concurring.
I join in the court’s judgment, but write separately to address appellee TJD.’s cross-issue challenging the factual sufficiency of the evidence supporting the jury’s finding that termination of her pa*517rental rights is in the best interest of her children.
The evidence in this case tracks the troubled journey of a young mother, T.D., whose terrible mistakes and personal failures culminated in her incarceration for criminal conduct and the loss of custody of her two young daughters, M.G.D. and B.L.D. The evidence also shows T.D.’s triumph over drug addiction and her hard-fought journey to reformation. Although the children endured great hardship as a result of their mother’s pre-incarceration conduct, and though they had found stability in a foster family that wanted to adopt them, the guardian ad litem opined that it was not in their best interest to terminate their mother’s parental rights. The trial judge agreed; the jury did not, electing in a 10-2 verdict to terminate T.D.’s parental rights. The record provides plenty of reason to question the factual sufficiency of the evidence supporting the jury’s finding, but, in the final analysis, this court is correct to conclude that a rational jury could have formed a firm conviction or belief that termination of T.D.’s parental rights is in her children’s best interest.
INVOLUNTARY TERMINATION of Parental Rights
Involuntary termination of parental rights is a serious matter, implicating fundamental constitutional rights. See Holick v. Smith, 685 S.W.2d 18, 18 (Tex.1985); In re G.M., 596 S.W.2d 846, 846 (Tex.1980). The natural parental right has been characterized as “essential,” a “basic civil right,” and “far more precious ... than property rights.” Stanley v. Illinois, 405 U.S. 645, 651, 92 S.Ct. 1208, 1212, 81 L.Ed.2d 551 (1972) (citations omitted); see Holick, 685 S.W.2d at 18. A termination decree is complete, final, and irrevocable, forever divesting that natural right as well as all legal rights (except for the child’s right to inherit). See Holick, 685 S.W.2d at 18. Because of these grave consequences, we must strictly scrutinize any decision that severs the natural tie between parent and child to be certain that children’s interests are best served and parental rights are protected. See id.
Evidence is factually insufficient to sustain termination of parental rights when a reasonable factfinder could not form a firm conviction or belief that termination is in the child’s best interest. See In re C.H., 89 S.W.3d 17, 25 (Tex.2002). In applying this heightened standard of review, we must be mindful of the respective constitutional roles of juries and reviewing courts. Id. Though we cannot be so rigorous that only fact findings established beyond a reasonable doubt withstand review, the evidence supporting the finding must be clear and convincing. Id. at 26.
The evidence in the record bears heavily on the children’s best interest at the time the Children’s Protective Services Division of the Texas Department of Protective and Regulatory Services (“CPS”) removed them from T.D.’s home. That evidence, though ample to show it was not in the children’s best interest to be in their mother’s care at the time of CPS intervention, is less convincing when it comes to showing termination of their mother’s parental rights was in their best interest nearly three years later, after she had shown signs of sustained reformation and rehabilitation.
The tragic story begins in October of 1994, when, at age sixteen, T.D. gave birth to M.G.D. The following year, in November of 1995, she had a second daughter, B.L.D.1 Several months after B.L.D.’s *518birth, T.D. took both children and moved from her mother’s home in Bacliff to Fort Worth to live with B.L.D.’s father and B.L.D.’s father’s parents. After about a month in Fort Worth, T.D. was asked to leave, so she returned to her mother’s home with M.G.D. and B.L.D. Over the next few years, T.D. and her children moved several other times.
In May of 1999, T.D., then twenty years old, was living with her daughters (then ages three and four) in a duplex in San Leon. According to her own testimony, at that time, T.D. was a “drug-addicted mother.” She made approximately $1,500 per week selling cocaine and was otherwise unemployed. When neighbors reported child neglect, CPS made a visit to T.D.’s home and found the house filthy and the children unkempt and infested with head lice. CPS removed the children for a short time, but returned them to T.D. when she signed a service plan. Shortly thereafter, T.D. and her daughters moved to Angelton, where T.D. continued to use and sell illegal drugs. In September of 1999, two police officers visited the Angle-ton residence in response to an anonymous child-neglect report. They arrested T.D. and charged her with possession of a controlled substance with intent to deliver and possession of marijuana. CPS removed the children and placed them in foster care. Three months later, T.D. was also charged with two counts of child endangerment. She was incarcerated from September of 1999 until August of 2000.
When CPS filed its first petition in September of 1999, its goal was family reunification. While this first suit was pending, CPS caseworkers visited T.D. in prison. In mid-September of 1999, when the duration of her incarceration was uncertain,2 T.D. signed a second service plan under which she was to help CPS locate the children’s fathers, develop a positive support system to assist her in raising her children, complete drug assessment through the Gulf Coast Recovery Center, remain drug and alcohol free, submit to random drug screening, complete individual counseling, undergo a psychological evaluation, and complete various training and educational courses. The second service plan also required T.D., upon her release from prison, to access resources that would facilitate her employment. In addition, she was to locate and maintain proper housing, and pay $150 per month to the Brazoria County Child Support Office.
During her incarceration, T.D. was not permitted visits with her children, but the service plan provided that she would have two one-hour sessions of supervised visitation with her children each month upon her release. T.D. sent letters to the children from prison, but had no other contact with them during the eleven-month period of her incarceration.
Meanwhile, the children were being moved from one foster home to another. Between September of 1999, when they were taken into CPS custody, and trial in May of 2002, M.G.D. and B.L.D. lived in three different foster homes. CPS removed them from the first foster home in *519February of 2000,3 and placed them in a second one. About a year later, CPS removed the girls again, this time placing them with foster parents who expressed a willingness to adopt them.
In early August of 2000, T.D. wrote a letter informing CPS that she would be paroled shortly. Just days before T.D.’s release from prison in late August of 2000, CPS served her with its first amended petition under which CPS sought termination of T.D.’s parental rights. The grounds for termination were the circumstances surrounding the children’s removal in September of 1999, and T.D.’s failure to fulfill her obligations under the second service plan. At trial, Terri Martin, CPS’s legal liaison and the girls’ caseworker from about April of 2000, until March of 2001 (hereafter, “CPS Caseworker”), testified the only requirement T.D. could have fulfilled while in prison was remaining drug and alcohol free, which she did.
The day after she was released from prison in August of 2000, T.D. called the CPS Caseworker to try to arrange a visit with her children. CPS, however, had asked the trial court to revoke T.D.’s visitation because she allegedly neither “visited nor maintained substantial contact” with the children for over a year. The CPS Caseworker acknowledged in her trial testimony that it was impossible for T.D. to have exercised visitation while she was incarcerated. Nevertheless, according to the CPS Caseworker, because CPS had decided to seek termination of T.D.’s parental rights, it did not want T.D. to see her children. The trial court granted CPS’s request and prohibited T.D. from seeing her daughters for the next few months.
In December of 2000, CPS’s termination suit was dismissed because of unspecified “procedural problems.” Later that same month, CPS filed a second termination suit in which it again sought termination of T.D.’s parental rights based on the same actions and omissions alleged in its first termination suit. With this filing, the timeline on the case started all over again so that, with an extension, CPS would have an additional eighteen months from December of 2000, to decide whether to reunite T.D. with her daughters.4
Though T.D.’s many previous efforts to get court-ordered visitation had proven unsuccessful, after the second case was filed, the trial court reinstated T.D.’s supervised visitation rights. From January of 2001 through the time of trial, T.D. had supervised visitation with M.G.D. and B.L.D. twice each month. Each visit was one-hour long and, at the foster mother’s request, took place in the Alvin CPS office, a four-hour drive for T.D., who had moved to Fairfield for employment. Despite the long commute, T.D. made nearly all of the visits, missing a few due to lack of transportation. During most visits the children greeted their mother with hugs and kisses, and the three then spent the hour visiting and playing games together. On some occasions, however, the children expressed a desire to not visit with T.D.
Shortly after acquiring supervised visitation rights, T.D. signed a third service plan. This plan went into effect in February of 2001, and required T.D. to complete essentially the same requirements contem*520plated by the second service plan, most of which were geared toward improving and acquiring parenting skills and providing a suitable home for her children. By June of 2001, T.D. had completed virtually all of the plan requirements.
The record shows after T.D. was released from prison, she: (1) found and maintained steady employment; (2) paid child support; (3) paid for health insurance for herself and her children; (4) completed parenting classes; (5) attended drug counseling; (6) paid outstanding fines and debts; (7) purchased a used car so that she could get to and from work; (8) paid a lawyer to help her keep her children; (9) took and passed 15 to 25 random drug tests conducted by both parole authorities and CPS; and (10) successfully completed her parole. Nonetheless, CPS would not reconsider its decision to seek termination of T.D.’s parental rights. At trial in May of 2002, the CPS Caseworker testified that CPS never reconsidered its decision to seek termination after August of 2000, notwithstanding T.D.’s substantial compliance with its requirements.
At trial, the CPS Caseworker and Mark Jones, the guardian ad litem appointed to protect the children’s interests (hereafter, “Guardian Ad Litem”),5 gave conflicting recommendations regarding the children’s best interest. The Guardian Ad Litem opposed termination of their mother’s parental rights and the CPS Caseworker favored it.
In the one-year period the CPS Caseworker was involved (April of 2000 until March of 2001), she talked to T.D. one time, by telephone. Though she observed some of the early visitation sessions, at the time of trial, the CPS Caseworker had not had any personal contact with T.D. or the children for over a year. She had not visited T.D.’s current home, nor does the record suggest CPS ever conducted a home study to gather information pertinent to the children's best interest or to evaluate T.D.’s progress after completion of services prescribed by the plan.
Necessary Findings for Termination of Parental Rights
To terminate parental rights, the trier of fact must find by clear and convincing evidence that: (1) the parent has engaged in one of the grounds for termination listed in section 161.001(1) of the Texas Family Code; and (2) termination is in the child’s best interest. Tex. FamGode § 161.001. Both elements of this test must be established, and termination statutes are strictly construed in favor of the parent. See Tex. Dep’t of Human Servs. v. Boyd, 727 S.W.2d 531, 533 (Tex.1987); see also Holick, 685 S.W.2d at 18.
There is no question that CPS met the first prong by adducing clear and convincing evidence of grounds for termination based on T.D.’s pre-incarceration conduct. The analysis for the best-interest prong is not so easy. If the trial had taken place three years earlier, the best-interest question would not be a difficult one for many of the reasons cited by the majority. But our analysis cannot be confined to that point in time. A termination decision must be based on the whole picture — one that includes an assessment of both the “then and there” and the “here and now.”
In T.D.’s case, the “then and there” is very different from the “here and now.” *521And this difference makes it all the more difficult to overcome the strong presumption that the best interest of M.G.D. and B.L.D. is served by preserving the parent-child relationship with their mother. See Wiley v. Spratlan, 543 S.W.2d 349, 352 (Tex.1976); In re D.T., 34 S.W.3d 625, 641 (Tex.App.-Fort Worth 2000, pet. denied). Thus, the jury’s verdict must be supported by clear and convincing evidence that termination of T.D.’s parental rights is in each child’s best interest — not just at the time CPS intervened, but at the time of trial.
The best interest of the child is often infused with the statutory grounds for termination under the first prong, but the best-interest determination must have a firm basis in facts standing apart from the offending behavior. In re W.C., 98 S.W.3d 753, 756-57 (Tex.App.-Fort Worth 2003, no pet.). For example, in Horvatich v. Tex. Dep’t of Protective and Regulatory Servs., although there was evidence that supported the trial court’s finding under section 161.001(1) of the Family Code, the Austin Court of Appeals found the evidence factually insufficient to support the trial court’s finding that termination of the mother’s parental rights was in the best interest of her three children. 78 S.W.3d 594, 601 (Tex.App.-Austin 2002, no pet.). There was evidence that the mother had exhibited poor parenting abilities and that she had failed to provide a safe and stable home environment for her children in the past. Id. at 600-01. However, the appellate court found this evidence would be insufficient even under a preponderance-of-the-evidence standard because there was no evidence regarding the children’s current conditions or future plans for the children, and the record suggested CPS did not adequately consider reunification or placement with relatives.6 Id. at 601-03.
The Horvatich court found the mother’s past failures amounted to some evidence that termination was in the children's best interest, but the record lacked evidence regarding the mother’s current parental abilities. Id. at 601-03. In the months leading up to trial, the mother had substantially complied with all of CPS’s prerequisites for reunification, though for a previous period of eighteen months she had no contact with her children and did not maintain regular contact with CPS.7 Id. At trial, the mother testified she loved her children and would be a better parent after her treatment. Id. at 598-99. She planned to finish her treatment two months after trial and then live in a halfway house for recovering addicts for six months. Id. She asked CPS to reunite her with her children after she completed treatment. Id. On remand, the appellate court recommended that the trial court fully consider the mother’s progress and efforts for reunification. Id. at 603.
Similarly, in In re W.C., the Fort Worth Court of Appeals reversed termination of a *522mother’s parental rights to her five children, finding the evidence was factually insufficient to support the jury’s best-interest determination. 98 S.W.3d at 765-66. CPS removed the children from the mother’s home because she continued to live with a man who abused the children. After CPS removed the children, the mother took parenting classes, attended counseling, completed random drug tests, maintained employment and her own apartment, and made significant progress in alleviating the cause of the children’s removal. Id. The appellate court found the best-interest decision was based on the mother’s past conduct and poor judgment, and did not take into account the uneontra-dicted evidence that she had done everything possible to be reunited with her children and could provide a safe, stable home for them. Id.
As these cases demonstrate, evidence of an abusive or neglectful parent’s turnaround can make a best-interest finding in favor of termination factually insufficient. Here, the record contains evidence which shows T.D.’s clear departure from past failures as well as conduct that is markedly different from the actions and omissions that led to CPS intervention. Courts should give evidence of parental turnaround coupled with material change in circumstances significant consideration in reviewing the factual sufficiency of the best-interest determination. This approach undergirds the important public policy favoring reunification of the family and honors the strong presumption that the best interest of a child is served by maintaining the parent-child relationship. See Wiley, 543 S.W.2d at 352; In re D.T., 34 S.W.3d at 641.
Evaluation or the Holley Factors and Other Relevant Considerations
In determining whether termination of parental rights is in a child’s best interest, courts consider a range of factors, commonly known as the Holley factors, which include the following:
(1) the desires of the child;
(2) the emotional and physical needs of the child now and in the future;
(3) the emotional and physical danger to the child now and in the future;
(4) the parental abilities of the individuals seeking custody;
(5) the programs available to assist these individuals to promote the best interest of the child;
(6) the plans for the child by these individuals or by the agency seeking custody;
(7) the stability of the home or proposed placement;
(8) the acts or omissions of the parent which may indicate that the existing parent-child relationship is not a proper one; and
(9) any excuse for the acts or omissions of the parent.
Holley v. Adams, 544 S.W.2d 367, 371-72 (Tex.1976). A factfinder is not required to consider all of the listed factors, nor is this list of relevant considerations exhaustive; other factors may be considered when appropriate. Holley, 544 S.W.2d at 372.
Present and Future Dangers to the Children
CPS provided ample evidence that the children were deprived and endangered in the months leading up to their initial placement in foster care. The evidence showed T.D. subjected her children to dangerous conditions and neglect. She failed to properly supervise and protect them. At times, she did not adequately provide for their healthcare and other basic needs. *523These were failures of omission. There is no evidence to suggest that T.D. was ever violent toward her children or that she took any action to injure them. CPS offered no evidence concerning present and future endangerment.
Present and Future Emotional and Physical Needs
Although a clinical psychologist evaluated the children and T.D., as ordered by the trial court, the psychologist did not testify as to his findings. CPS presented no expert testimony at all. The Guardian Ad Litem provided a layman’s assessment of the children’s bonding with their current foster mother and T.D. He testified that the children get along well with their foster mother and her two sons. After observing T.D. with M.G.D. and B.L.D., the Guardian Ad Litem concluded she seemed to lack an emotional connection with the children, noting T.D. appeared to be more like a family friend than a mother.
The Guardian Ad Litem attributed T.D.’s seeming lack of emotional connection with her daughters, at least in part, to the strained conditions under which the monthly visits occurred. Over half of the visits took place in the waiting room of the Alvin CPS office. The CPS Caseworker acknowledged that this was not an appropriate place for visitation. Lee Rieger, another CPS caseworker, likewise acknowledged that this meeting place did not provide “good quality visit time.” According to trial testimony, various CPS workers came and went during the visits and those who stayed in the room talked among themselves; on one visit, a worker reportedly walked through the room singing a song. If the talking and singing did not provide enough distraction, another CPS worker observing the session took notes in plain view of T.D. and her children. The Guardian Ad Litem testified that both the children and T.D. knew they were being observed, and these conditions adversely impacted the quality of the visits.
During some visits, CPS workers allowed one or both of the children to leave the visitation area. T.D. testified that the foster mother was in a nearby office during one visit and was present on several other occasions during T.D.’s visitation, which T.D. viewed as an impediment to her interaction with the girls. The visitation setting offered little opportunity for any meaningful time for T.D. and her daughters. CPS never presented an opportunity for T.D. to experience a private visit with her children, for example, in a separate room which enabled CPS to monitor the visit through a window.
The majority states “T.D. is critical of CPS’s efforts to help her be a better parent.” Actually, T.D. is critical of CPS’s efforts to thwart contact with her children. She testified that initially she was not given the correct address to write them while she was in prison and that after she was released, CPS blocked her many efforts to obtain visitation with the girls, and instead urged her to “walk away.” The Guardian Ad Litem made similar observations, noting that CPS had not been fair to T.D. He testified that it would have been beneficial for T.D. to have been given more visitation with the children and that despite his request, CPS had not provided more visits.
Despite T.D.’s efforts to maintain and enhance contact with her daughters, the separation from the girls for the eleven-month period she was incarcerated (and during the three-month period her visitation was blocked after her release from prison) undoubtedly took its toll on the mother-daughter relationships. Still, T.D. worked to restore what had been lost, making the most of what little time and *524opportunity she was given to be with her children. Though her focus was not on gift-giving, in her first post-incarceration visit with her daughters, T.D. brought the girls an assortment of toys and candy. She testified it was the first time she could really afford to buy something for her children. Before Christmas of 2001, T.D. asked the children what they wanted. To T.D.’s surprise, the children warned she was not to buy them gifts because their foster mother said they did not need them. This exchange gave T.D. the impression that if she bought gifts for the children, they would not be allowed to keep the gifts.
Despite the hardships M.G.D. and B.L.D. have endured, there was no evidence at trial to suggest they have extraordinary emotional or physical needs. Aside from the testimony of the Guardian Ad Litem, who ultimately opposed termination of T.D.’s parental rights, and a stipulation that the children’s needs were being met in foster care, CPS offered no evidence, direct or indirect, concerning the children’s present and future emotional or physical needs.
Acts or Omissions of the Parent
In assessing the acts or omissions of the parent, it is significant to note that T.D.’s criminal conduct and other problem behaviors that led to her children being removed from the home were rooted in severe substance abuse. When questioned about why she had used drugs in the first place, T.D. stated that she could not blame anyone except herself, but she explained that it was partially attributable to “the way [she] was raised” and that it had “to do with feeling rejected” and unloved as a child. T.D. recounted her own painful childhood that was marked by an early exposure to drugs, violence, neglect, and abuse in the home.
T.D.’s parents divorced before her second birthday. Her father was in prison or on the road as a truck driver most of her childhood. T.D. was molested as a child, and CPS took her from her home when she was in the second grade. She lived in a foster home for several months before being returned to her mother. T.D. testified that, as a child, she witnessed her mother being verbally and physically abused by various boyfriends. When T.D. was thirteen years old, her mother “kicked her out of the house.” T.D. lived in an abandoned house until her mother allowed her to return home. After that, T.D.’s mother “kicked her out of the house about every two days.” Though she eventually received a high school equivalency certificate, T.D. did not attend school beyond the seventh grade.
T.D. began to abuse drugs during her adolescent years. Her mother openly used marijuana and abused alcohol throughout T.D.’s childhood, and she allowed T.D. to use marijuana in their home. T.D. explained that her drug use was a progressive addiction and that “once [she] got started, [she] couldn’t stop.” She began using drugs at an early age and she became addicted to them. Eventually, she began to sell drugs. She abused drugs to the point of “putting drugs before her children.” It was not until she was incarcerated that T.D. was finally able to effectively address her substance-abuse problems and at last become drug-free.
Though T.D.’s childhood and adolescence were plagued by abuse, neglect, and abandonment at the hands of her parents, that dreadful life experience does not excuse her neglect of her own children. T.D. herself admits this. But T.D.’s story is a compelling one, and her many efforts to overcome her tragic past inform the fact-finder’s evaluation of her ability to sustain reform and succeed in her future plans.
*525If a wayward parent has taken no action to address an addiction or to correct other problem behaviors that led to CPS intervention — through, for example, counseling, therapy, and education — then a rational factfinder could reasonably conclude that the parent will not likely be successful in modifying those behaviors and effecting positive change for the future. If no meaningful action is taken, the parent will not be equipped with the necessary tools to guard against reoccurences. Conversely, a rational factfinder would likely conclude the prospects for success are much more promising for a parent who gets help, takes advantage of professional services, addresses problem behaviors and deficiencies in parenting, and demonstrates material change. The record shows T.D. did just that.
Referring to her drug abuse and the many inappropriate behaviors that went with it, T.D. candidly admitted that the things she did before her incarceration were “totally and completely wrong.” In her words, she does not “do the same things” anymore. Though she testified that she no longer had a substance-abuse problem, T.D. acknowledged she would always be an addict, demonstrating an important awareness of her vulnerability to addiction. She explained, however, that there is a difference between being an addict and using drugs. T.D. testified that she participated in a twelve-step recovery program and explained that it is an “ongoing thing” for “day-to-day life ... one day at a time.” At trial, even the CPS Caseworker acknowledged that T.D. had effectively addressed her drug addiction — the root of the problems that led to her incarceration and the loss of custody of her children.
Parental Abilities and Available Programs
The evidence shows that before her incarceration, T.D. engaged in conduct that demonstrated a lack of fundamental parenting abilities. In evaluating parental abilities, however, the factfinder must consider not only the skills lacking at the time of CPS’s intervention but also any skills acquired in the interim. A parent’s material improvement in the areas identified as deficient is some indication of the prospects for success in the future. Consequently, a just evaluation of T.D.’s parental abilities cannot be based solely on what transpired before September of 1999, but also must include a fair consideration of what happened in the nearly three years since CPS intervened.
After she was released from prison, T.D. took many constructive steps to acquire the parenting skills she desperately needed to protect her children and provide them with a healthy, safe, and stable environment. The first step in this process was her acknowledgment that she was ill-equipped and unprepared for the task and had made many terrible mistakes in the past. After addressing her substance-abuse problem, T.D. turned her attention to the deficiencies in her parenting abilities. She met, and in some respects, exceeded the requirements of the third service plan.
The record suggests T.D. has learned valuable parenting skills and life lessons which she is eager to put into practice. She received a certificate for completing a parenting course that taught how to identify the cause of past parental failure and how to implement various parenting techniques. When asked whether she was concerned that the cycle of neglect and violence she had experienced as a child was being continued in her children, T.D. was transparent in her response, candidly explaining that there can be no guarantees, but expressing a firm commitment to uti*526lize her newly acquired knowledge if given the opportunity. She stated:
I can say, for one, that I know that I was given an opportunity to learn things that my mother never was. I kind of used this intervention to better myself. I can’t honestly guarantee that it won’t be a repeated cycle, but I can do my best to not. I know that the key to making mistakes is to learn from it and to not make the same mistakes over and over. I know that I’ve never wished any of what’s happened to happen to my kids because I’ve always wanted to do it differently than my mother raised me; and in turn, I[did] the same trying not to, but now that I have better knowledge of how to do it correctly and how not to do it, I feel that given a chance to — how do I say — use that knowledge, that I could only do better and better by learning from what I’ve already learned and keep learning.
T.D. also expressed her desire to continue receiving services “if allowed.” The majority belittles T.D.’s recognition that additional services would be helpful, stating, inaccurately, that T.D. “complains the agency should have spent more time and money teaching her how to be a better parent.” T.D.’s yearning to further develop her parental skills beyond the service plan’s requirements should be applauded, not condemned. If anything, T.D.’s desire for additional help evinces a determination to further her progress. Though the majority questions the efficacy of parenting courses to achieve this end, classroom instruction and practical guidance on effective and appropriate parenting strategies can be very useful to someone like T.D., who had little opportunity to observe or acquire basic parenting skills from her own family experience.
T.D.’s parenting classes and counseling totaled at least 96 hours of instruction. She attended three times a week, four hours a day, for a five or six month period. She completed the programs successfully one month before the target completion date. T.D. received her completion certificates for the courses and counseling and even continued to see counselors from time to time on her own, after she finished the course work. The evidence shows CPS can and will continue to provide additional services and instruction to a parent even after a failed effort to terminate parental rights.
The majority also states that once CPS decided to seek termination of T.D.’s parental rights, “it is not clear why the agency should have poured greater resources into bringing about the opposite result.” The answer lies in the longstanding public policy in favor of maintaining the natural parent-child relationship. See Wiley, 543 S.W.2d at 352. If CPS ignores this fundamental policy and does not adequately consider reunification or other viable alternatives to termination, this fact can be considered in evaluating the factual sufficiency of the evidence. See Horvatich, 78 S.W.3d at 602.
Moreover, common sense suggests that after investing so much time, money, and resources to effect change in behavior through therapy, counseling, and education, CPS would want to measure the results of its efforts. In this case, CPS dutifully prescribed the requisite services, but showed little interest in the fruits of its labor, as exemplified by the CPS Caseworker’s total lack of contact with T.D. for more than a year before trial, her failure to visit T.D.’s home, and her failure to assess change and check for improvement after T.D.’s completion of extensive services. In short, CPS had no interest in the “here and now,” focusing instead on T.D.’s pre-September 1999 conduct.
*527What is not clear is why the CPS Caseworker did not conduct any follow-up to assess T.D.’s progress before recommending termination of her parental rights. Though she testified that CPS never reconsidered its August 2000 decision to seek termination, the CPS Caseworker offered no explanation for the apparent lack of interest in T.D.’s progress. Why counsel neglectful or abusive parents about the causes of parental failure if not to effect changes in their behavior? Why offer parent training and instruction if not to help them develop the skills to successfully parent their children? And why do any of it if there is no follow-up to check for progress?
Plans for the Children and Stability of the Home
In describing her future plans for the children, T.D. explained that she planned to take care of them “in a way that [she] couldn’t before,” using the knowledge and skills she has gained from her educational courses, counseling, and therapy. T.D. pointed out that she had the means to provide for her children, noting that she had a place to live8, food to eat, a job, and a car and could take care of her daughters. T.D. also testified about her plans for child care while she is working, explaining that she planned to utilize services her current employer provides to assist employees in locating day care for children.
T.D. has been paying child support as well as health insurance costs for herself and her children since she started her current job at a construction company. At the time of trial, she was working seven twelve-hour shifts each week. She looked forward to starting a two-year assignment with her company in Rosenberg. Transferring to this location would put her in closer proximity to a cousin who is a homemaker and mother and who could help T.D. with the girls. Though some of her immediate family members have criminal records, T.D. pointed out that she has many family members without criminal records who could help her with child care and serve as a family support system.
Though T.D. failed to provide a stable home life for her daughters in the past, the material changes in her home, habits, lifestyle, finances, and employment status strongly suggest that she would be able to provide far more stability in the future. The trial judge noted on the record that T.D. already had achieved some of the stability necessary to parent her children.
Compliance with Service Plans
A parent’s compliance with service plans can be useful in determining the children’s best interest. See In re W.C., 98 S.W.3d at 765. Although T.D. did not comply with the first service plan, after she was released from prison, she demonstrated a marked turnaround from her initial noncompliance by taking advantage of several programs, services, and resources that had been unavailable to her during her incarceration. At the time of trial, she demonstrated nearly complete compliance with CPS service plans.
Although T.D.’s compliance with the CPS service plans does not insulate her from a finding that termination of her parental rights is in the children’s best interest, it is a factor that takes on added significance when coupled with material *528change. It is hardly debatable that T.D. effected material change in many aspects of her life in the nearly three-year period at issue here.
The CPS Caseworker’s Recommendation
Despite the lack of a home study and notwithstanding her lengthy lack of contact with T.D. prior to trial, the CPS Caseworker testified she believed termination was in the children’s best interest. The reason the CPS Caseworker gave for her recommendation was that the children were secure, thriving in foster care, being nurtured, and doing well. This best-interest conclusion bespeaks a fundamental misunderstanding of the best-interest standard for termination of parental rights.
The best-interest standard does not permit termination of parental rights merely because a child might be better off living elsewhere. See In re D.M., 58 S.W.3d 801, 814 (Tex.App.-Fort Worth 2001, no pet.) Nor should termination be used as a means of re-allocating children to better and more prosperous parents. Id. Though the children have had a successful experience in their current (third) foster home, this should not overshadow the principal goal of reunification of a natural parent with her children, particularly when the parent has completed the services CPS prescribed and taken constructive steps to remedy the deficiencies that led to CPS intervention.
The Guardian Ad Litem’s Recommendation
At trial, the Guardian Ad Litem testified that termination of T.D.’s parental rights would not be in her daughters’ best interest. Synthesizing this assessment with the CPS Caseworker’s must begin with recognition that any assessment is only as strong as the investigation that supports it. Because the CPS Caseworker and the Guardian Ad Litem conducted very different levels of investigation, they had very different knowledge of relevant events occurring over the long period preceding trial. This may explain why their assessments and recommendations stood in sharp contrast.
By her own admission, the CPS Caseworker conducted no investigation of T.D.’s current situation, and so had no knowledge of the “here and now.” The Guardian Ad Litem undertook a much more comprehensive factual analysis before making his recommendation. Rather than focus exclusively on T.D.’s conduct that led to CPS intervention, the Guardian Ad Litem took a broader view and based his evaluation on T.D.’s behavior both before and after those events. In doing so, he concluded termination of T.D.’s parental rights was not in the children’s best interest.
The Guardian Ad Litem expressed his concerns not only that CPS had failed to conduct a home study but also about the way CPS had conducted the visitation sessions between T.D. and her children. He testified that T.D. had not been given a “fair shot.” Though the Guardian Ad Li-tem did not believe T.D.’s parental rights should be terminated, he recommended that the children remain in their current foster home for the time being.
The Trial Judge’s Concerns and Judgment Notwithstanding the Verdict
The trial judge also disagreed with the CPS Caseworker’s conclusion that termination was in the best interest of the children. In announcing his decision to disregard the jury’s best-interest finding, the trial judge showed he, too, was troubled by the lack of evidence as to the children’s *529present and future best interest. He noted that CPS decided to terminate regardless of T.D.’s completion of service plans, and that after her release from prison, T.D. only exemplified perseverance. The trial court also noted T.D. had completed drug therapy, counseling, and parenting classes, and achieved some of the stability called for in the service plans.9 For these reasons, the trial judge decided to disregard the jury’s best-interest finding. He appointed CPS as the permanent managing conservator so CPS could continue to help T.D. and the children until T.D. is ready to have custody.
The Applicable Standard of Review
Ordinarily, courts must give substantial deference to the jury’s findings of fact. However, in some circumstances the United States Constitution requires that courts apply an independent, de novo standard of review. Courts apply this de novo standard of review to fact findings regarding the constitutionally mandated requirements that actual malice be proven by clear and convincing evidence in certain defamation cases and that punitive-damage awards satisfy federal due process requirements. See Cooper Indus., Inc. v. Leatherman Tool Group, Inc., 532 U.S. 424, 436, 121 S.Ct. 1678, 1685, 149 L.Ed.2d 674 (2001) (punitive damages); Bose Corp. v. Consumers Union of U.S., Inc., 466 U.S. 485, 510-511, 104 S.Ct. 1949, 1964-65, 80 L.Ed.2d 502 (1984) (actual malice). Neither the United States Supreme Court nor the Texas Supreme Court has addressed whether appellate courts should review de novo the constitutional requirement that termination of parental rights be proven by clear and convincing evidence, but some state courts have so held. See Santosky v. Kramer, 455 U.S. 745, 769-70, 102 S.Ct. 1388, 1402-03, 71 L.Ed.2d 599 (1982) (holding federal due process requires the State to prove allegations in parental-termination cases by at least clear-and-convincing evidence but not addressing the standard of appellate review); Thompson v. Thompson, 110 Idaho 93, 714 P.2d 62, 64-65 (1986) (discussing conflict among state courts as to whether de novo review is required and holding it is not); In re S.B.C., 64 P.3d 1080, 1081-83 (Ok.2002) (holding that federal constitution requires de novo review of jury’s fact findings in parental-termination cases); In re C.H., 89 S.W.3d at 29 (Hect, J., concurring) (noting that federal constitution may require de novo review in parental-termination cases but that court did not need to reach that issue because it did not affect the outcome in that case). T.D. has not asserted that the United States Constitution requires a de novo standard of review, and this court is bound to apply the standard of review announced in In re C.H.
Nonetheless, appellate courts have a responsibility to ensure that parents and children receive fair, consistent, and expeditious appellate review in parental-termination cases. See In re J.F.C., 96 S.W.3d 256, 298 (Tex.2002) (Hankinson, J., dissenting). The current state of the law makes it difficult for courts to adequately discharge this responsibility. At present, in Texas parental-termination cases there is uncertainty as to whether the federal constitution requires a de novo standard of appellate review. Furthermore, the common-law standards of review recently announced by the Texas Supreme Court for reviewing the sufficiency of the evidence blur the distinction between legal and fac*530tual sufficiency. See In re J.R.K, 104 S.W.3d 341, 342-43 (Tex.App.-Dallas, 2003, no pet. h.) (noting that the standards of review for legal and factual sufficiency in parental-termination cases are “essentially the same”). This point is keenly apparent in the Texas Supreme Court’s explanation of the fragile distinction between these two standards of review:
The distinction between legal and factual sufficiency when the burden of proof is clear and convincing evidence may be a fine one in some cases, but there is a distinction in how the evidence is reviewed. In a legal sufficiency review, a court should look at all the evidence in the light most favorable to the finding to determine whether a reasonable trier of fact could have formed a firm belief or conviction that its finding was true. To give appropriate deference to the factfinder’s conclusions and the role of a court conducting a legal sufficiency review, looking at the evidence in the fight most favorable to the judgment means that a reviewing court must assume that the factfinder resolved disputed facts in favor of its finding if a reasonable factfinder could do so. A corollary to this requirement is that a court should disregard all evidence that a reasonable factfinder could have disbelieved or found to have been incredible. This does not mean that a court must disregard all evidence that does not support the finding. Disregarding undisputed facts that do not support the finding could skew the analysis of whether there is clear and convincing evidence.
If, after conducting its legal sufficiency review of the record evidence, a court determines that no reasonable factfinder could form a firm belief or conviction that the matter that must be proven is true, then that court must conclude that the evidence is legally insufficient ...
In a factual sufficiency review, as we explained in In re C.H., a court of appeals must give due consideration to evidence that the factfinder could reasonably have found to be clear and convincing. We also explained in that opinion that the inquiry must be “whether the evidence is such that a factfinder could reasonably form a firm belief or conviction about the truth of the State’s allegations.” A court of appeals should consider whether disputed evidence is such that a reasonable fact-finder could not have resolved that disputed evidence in favor of its finding. If, in light of the entire record, the disputed evidence that a reasonable factfinder could not have credited in favor of the finding is so significant that a factfinder could not reasonably have formed a firm belief or conviction, then the evidence is factually insufficient.
In re J.F.C., 96 S.W.3d at 267 (footnotes omitted). Although some may question whether the standard of review in In re C.H. sufficiently protects the fundamental parental rights at issue in these termination cases, we must apply this standard in determining T.D.’s cross-issue.
Assessing Factual Sufficiency
In determining whether factually sufficient evidence supports the jury’s verdict, we must consider all of the evidence in the case, evidence regarding both T.D.’s pre-September 1999 conduct and her subsequent reformation, giving due consideration to her current situation. See In re C.H., 89 S.W.3d at 28; Maritime Overseas Corp. v. Ellis, 971 S.W.2d 402, 406-07 (Tex.1998). The evidence of the latter militates against finding that termination is in the children’s best interest. See Hendricks v. Curry, 401 S.W.2d 796, 800 (Tex.1966) (stating that termination of parental rights should not be based solely on conditions that existed in the distant past but no *531longer exist); In re R.R.F., 846 S.W.2d 65, 68-69 (Tex.App.-Corpus Christi 1992, writ denied) (stating same principle as Hendricks and holding that conduct five years before trial was too remote to support termination, where there was no proof that this conduct created a present or future threat to the children); Wetzel v. Wetzel, 715 S.W.2d 387, 389-91 (Tex.App.-Dallas 1986, no writ) (holding that, in termination suit, acts done in the distant past, in this case approximately three to four years before trial, are insufficient to support termination unless the evidence proves a present or future danger to the child); Johnson v. Jefferson County Child Welfare Unit, 557 S.W.2d 569, 571 (Tex.Civ.App.-Beaumont 1977, no writ) (holding that “distant past” rule of Hendricks applied to conduct three years before trial). Unquestionably, T.D.’s pre-incarceration conduct put any opportunity for reunification with her children at serious risk, but her post-incarceration conduct was shown to be dramatically different. She is no longer a “drug-addicted mother” who sells cocaine for a living; she is drug-free. She is no longer incarcerated and she has successfully completed parole. She is no longer unemployed, but is working a steady job and providing financial support and health benefits for her children. She has rid herself of destructive relationships, completed parenting courses, benefitted from a twelve-step recovery program, attended counseling, and continues with therapy. She has effected material changes in her life and circumstances.
To sustain the jury’s finding, this court must conclude that notwithstanding all of T.D.’s reformative measures, a rational factfinder could form a firm conviction or belief that termination of her parental rights was nonetheless in her children’s best interest. See In re C. H., 89 S.W.3d at 28. The Guardian Ad Litem could not do so. Neither could the trial judge. Given the strength of the evidence of T.D.’s turnaround and reformation, their conclusion that termination of T.D.’s parental rights was not in her children’s best interest is understandable. That does not mean, however, that a rational jury hearing the same evidence could not reach the opposite conclusion.
The jury heard the evidence of T.D.’s conduct that led to CPS intervention as well as the evidence of her turnaround and reformation. Under current law, an appellate court cannot reweigh the evidence on appeal, nor can it substitute its judgment of the children’s best interest for the considered judgment of the jury that decided the matter. The totality of the evidence is such that a reasonable factfinder could form a firm conviction or belief that termination of T.D.’s parental rights is in the children’s best interest notwithstanding her turnaround and signs of sustained reformation. Thus, under the applicable standard of review, the evidence is factually sufficient to support the jury’s best-interest finding.
. T.D.’s two daughters have different fathers, and T.D. was never married to either man. The fathers' parental rights were terminated *518in the trial court and are not relevant to this appeal.
. CPS initially tried to place the girls with T.D.’s family members. CPS could not place the children with T.D.’s mother because she fell ill in September of 1999, and died in November of 1999. From early 2000, until a few months before T.D.’s release from prison in August of 2000, CPS tried to place the children with T.D.’s aunt. Placement with the aunt fell through when the aunt became reluctant to take the girls into her home. CPS was also concerned because the aunt’s boyfriend was a felon.
. The first foster mother testified B.L.D. sexually acted out several times while in her care. Nothing in the record indicates T.D. caused this behavior. The trial court excluded testimony that this acting out is why the children were moved to a second foster home.
. CPS had one year from the date it filed its first petition to decide whether to reunite the family or pursue other plans for the children. A one-time extension for six additional months was available with court approval. See Tex Fam.Code § 263.401(b).
. At the time of trial, the Guardian Ad Litem had been practicing law for nearly ten years, with sixty percent of his practice concentrated in family law. The Guardian Ad Litem was appointed under section 107.001 of the Texas Family Code to represent the best interest of the children and provide the court with impartial recommendations. See Tex Fam.Code § 107.001.
. Only the children's former caseworker, who worked on the case until the children were removed from their mother, and the assigned conservatorship worker, whose work on the case had ended ten months before trial, testified about the condition of the children. Horvatich v. Tex. Dep't of Protective and Regulatory Servs., 78 S.W.3d 594, 599 (Tex.App.-Austin 2002, no pet.). There was no testimony on the children’s condition for at least the ten months preceding trial or on whether they were being considered for adoption. Id.
. The mother had completed a psychological evaluation, parenting and employment classes, a family-strengthening program, visits, individual counseling, and drug and alcohol assessments. Id. at 603. The record also contained evidence of her negative drug-test results. Id. Child-support payments were the sole requirement with which she failed to comply. Id.
. T.D. stated that with more time, she could obtain better housing. At the time of trial, T.D. lived in a trailer home, which she leased for $125 per week. Though she lived there alone, it could sleep five people. She planned to look for a bigger place, but explained that, because she did not currently have custody of her children, there was no need for a larger home at that time.
. In addition, the trial judge made two observations about the Guardian Ad Litem’s testimony. First, T.D. was not given a fair opportunity to complete the services CPS required under the reunification plan. Second, the Guardian Ad Litem did not believe termination of T.D.’s parental rights would serve the children’s best interest.