This appeal presents once again the issue of paternity by estoppel. See Dews v. Dews, 632 A.2d 1160, 1166-69 (D.C.1993). Appellant, K.A.T., contests an order for child support in favor of appellee, C.A.B., and her minor daughter, R., asserting that the trial judge erred in holding that appellant was equitably estopped from denying that he was the father of the child, and thus obligated to pay child support. Because our decisions are incompatible with a conclusion of equitable estoppel on the facts of this case, we are constrained to reverse.1
I.
C.A.B. (the mother) became pregnant with R. (the child) in 1974. For purposes of this litigation, the father is conceded to be D.P., a man the mother had known since her own childhood but who quickly lost interest in a paternal relationship with the child. He saw the child only once after her birth in March of 1975, and neither mother nor child has seen him since or has knowledge of his whereabouts. On the other hand, the record does not reveal that any efforts have been made to locate D.P.
Several months before the birth, the mother resumed contact with appellant (K.A.T.), with whom she had had an earlier friendship. The couple began dating and continued doing so after the child’s birth. From the beginning the child called K.A.T. “daddy,” and there was testimony, credited by the trial judge, that K.A.T. referred to her as his daughter when in the company of others. Although there was no testimony that the mother and K.A.T. lived together or shared the expenses of raising the child during the *571first three years of the child’s life,2 K.A.T. did participate regularly in outings with the child such as picnics and attending movies, and he sometimes bought the child clothing and toys. The child was never told dining this period that he was not her father.
In 1978 K.A.T. moved to California where he remained until approximately April of 1981. While there he remained in touch with the child by telephone, letters, and gifts sent to her by mail. At some point the mother visited him in California and they discussed marriage. According to testimony which the trial judge credited, K.A.T. declared his desire, upon marriage, to treat the child as his daughter and agreed to “do whatever he [could] to help to support” her, including setting up a trust fund for her college education. It does not appear that the fund was ever established. The couple were married in September 1981. At about that time K.A.T. signed an Acknowledgement of Parentage filed with the Government of the District of-Columbia Vital Records Division, declaring himself to be the child’s father. The record reveals no steps, however, toward adoption of the child by K.A.T.
The couple began having marital difficulties in 1982 and separated in November 1983. During this two year period, the child lived partly with the couple and partly with the maternal grandmother. In 1983 the child overheard an argument between the mother and K.A.T., during which she learned for the first time that K.A.T. was not her biological father. Despite the separation, from 1982 through 1985 the couple filed joint federal income tax returns claiming the child as a dependent.3 They divorced in early 1986, executing a property settlement agreement under which K.A.T. was to pay the mother a lump sum of $8,000 in return for the deed to the marital home acquired during the marriage; but the agreement made no provision for child support. Following both the separation and the divorce, K.A.T. continued to buy clothes and other gifts for the child, visited her perhaps twice a month, and maintained contact by telephone. She continued to regard him as the only father she had known.
The trial judge, although not having the benefit of our opinion in Dews, supra, rendered a thoughtful written decision in which he found, among other things, that K.A.T. had misled the child into “believ[ing] that he was her natural father by a conscious course of conduct for thirteen years,” and that it would be inequitable for him to “eontinue[ ] to enjoy the benefits of an established paternal relationship with a thirteen year old child ... without assuming the concomitant financial obligation of such a relationship.”
II.
This case does not involve the obligations of one standing in loco parentis. Our decisions make clear that that status, itself involving “a somewhat nebulous legal relationship,” Cooley v. Washington, 136 A.2d 583, 585 (D.C.1957), terminates upon divorce and even upon separation of the husband and wife, “unless the party standing in loco parentis to the child means that it should continue.” Jackson v. Jackson, 278 A.2d 114, 115 (D.C.1971) (footnote omitted). No one contends in this case that K.A.T. intended the “strictly temporary” obligations, id. (citation omitted), he assumed toward the child during the parties’ brief marriage before separating to continue after dissolution.
Rather, this case involves the kindred equitable notion of paternity by estoppel. In general, the doctrine of equitable estoppel has been applied with great reluctance in the paternity context. Revisiting the subject in Dews, supra, this court canvassed the decisions and concluded that, while “[m]any appellate courts, including this one, have considered whether the doctrine of equitable estoppel might be applied to prevent a supposed father from denying paternity, ... none that we have found has affirmed a trial *572court decision to that effect.” 632 A.2d at 1167 (citing cases). Although appellee is able to point to somewhat greater receptivity toward the doctrine recently, see, e.g., Pietros v. Pietros, 638 A.2d 545 (R.I.1994), nothing like a trend can be identified toward imposing the statutory obligations of parenthood on a nonparent by way of equitable considerations. The reasons for the prevailing “caution,” Miller v. Miller, 97 N. J. 154, 478 A.2d 351, 358 (1984), in applying equitable estop-pel in this context are essentially twofold. The first is that statutory law, in the District of Columbia as elsewhere, imposes the primary duty of child support on the natural parents, see, e.g., Prager v. Smith, 195 A.2d 257, 259 (D.C.1963); D.C.Code § 16-916 (1989); moreover, the legislature has long provided an express means — adoption—by which a non-biological parent may acquire the formal and permanent status of parent. E.g., Fuller v. Fuller, 247 A.2d 767, 769 (D.C.1968), petition denied, 135 U.S.App.D.C. 353, 418 F.2d 1189 (1969); D.C.Code § 16-302 (1989). The second reason is the deep practical concern that extending the in loco parentis relationship beyond dissolution of marriage would “discourage the voluntary support that is so often provided by stepparents.” Dews, 632 A.2d at 1169. - As the Supreme Court of New Jersey stated in Miller, supra, in holding that “the development of ‘emotional bonding’ ... is not sufficient to invoke the doctrine of equitable estoppel in stepparent cases”:
[T]o hold otherwise would create enormous policy difficulties. A stepparent who tried to create a warm family atmosphere with his or her stepchildren would be penalized by being forced to pay support for them in the event of a divorce. At the same time, a stepparent who refused to have anything to do with his or her stepchildren beyond supporting them would be rewarded by not paying support in the event of a divorce.
478 A.2d at 358.
These considerations led us in Dews to observe that, while “[t]here may well be circumstances in which it is appropriate to impose on a stepparent the obligation to support a stepchild by invoking the doctrine of equitable estoppel[, tjhere should ... be no doubt about when that doctrine will be applied....” 632 A.2d at 1169. K.A.T. contends that we can insure rejection of the doctrine in doubtful cases only by adopting the rule established in Miller, supra, that “[t]he stepparent must take positive action interfering with the natural parent’s support obligation to be bound” by equitable estop-pel. 478 A.2d at 359 (emphasis added). We think it unnecessary to require such a limitation in this case, for even under a less restrictive approach taking into account all of the circumstances of the case, our decisions preclude application of the doctrine to these facts.4
“Equitable estoppel to deny a duty to support [a child] can be invoked only upon a showing that there was an express or implied misrepresentation of fact inducing another to alter his position to his prejudice.” Fuller, 247 A.2d at 769 (footnote omitted). We explained in Fuller that the “other” in question is the child; “any misrepresentation ... can have relevance only if made to [him or] her.” Id. Although the notion of an infant or pre-adolescent child being induced “to alter his position” is somewhat artificial, we reiterated in Dews, supra, that the party claiming equitable estoppel in the paternity context must meet the threefold prerequisites of that doctrine generally: misrepresentation, as well as reliance and resulting “change [of] position” by (or at least detriment to) the child. 632 A.2d at 1167 (quoting Parker v. Sager, 85 U.S.App.D.C. 4, 8, 174 F.2d 657, 661 (1949)).
Considering first the element of misrepresentation, the trial judge found that K.A.T. had “encouraged, or at least-permitted, [the child] to believe that he was her natural father by a conscious course of conduct for thirteen years.” If by this the judge meant only that K.A.T. never told the child he was not her natural father, the judge is correct— although the parties agree that in 1983, when the child was eight years old, she accidentally *573learned the truth that K.A.T. was not her biological father. Fuller, however, calls into substantial question the factual underpinnings of the judge’s finding of misrepresentation. The judge found, among other things, that K.A.T. had allowed the child “to call him ‘daddy’ virtually from the time she learned to talk”; that he “always referred to her as ‘his daughter’ when she was with him in the company of others”; and that at the time of the marriage he executed an Acknowledgement of Parentage with the Vital Records Division. But Fuller held that neither “the fact that the child referred to [the putative father] as ‘Daddy’ ” nor that the latter “misrepresented to the Bureau of Vital Statistics and to the public generally that he was the natural father” evinced “such a misrepresentation ... toward the child as would warrant the application of equitable estoppel in the circumstances of this case.” Fuller, 247 A.2d at 769 (emphasis added). It is questionable whether the remaining “circumstances” of this case differ enough from Fuller5 to make this the unambiguous case of misrepresentation that that case and Dews require.6
We need not rest our decision on this ground, however, because we are not convinced that the other prerequisites of reliance and prejudice (detriment) have been met in this case. The trial judge found that the child “was adamant that [K.A.T.] was the only ‘daddy’ she had ever known.” But Fuller rejected a claim of reliance based on similar evidence of emotional bonding, stating that:
[n]o matter what the beliefs of the parties [including the child] may have been, [the putative father] did not, by taking the child into the family circle, effect an adoption of her, thereby imposing upon him the attendant continuing obligation of support and entitling the child to certain other legal rights.
Id. Fuller thus anticipated a similar conclusion reached by the Supreme Court of New Jersey in Miller, supra:
It is undisputed that Jay [the putative father], while living with Gladys [the mother], developed a loving relationship with the girls, and that the girls relied on him for emotional support. However, no court has ever applied equitable estoppel to force a husband to support the children of his divorced spouse merely because he developed a close relationship with the children, nurtured them into a family unit with himself as the father, and had the children call him “daddy.” We decline to be the first to set such a precedent.
Miller, 478 A.2d at 858.
Fuller concluded that the child’s emotional dependency on the putative father was not “prejudice[ ]” sufficient to invoke equitable estoppel, Fuller, 247 A.2d at 769, a conclusion we perforee reach here as well. That leaves us with the trial judge’s determination that the child would suffer financial detriment if K.A.T. were “permitted, in essence, to disown her after a thirteen year father-daughter relationship.” But this conclusion also suffers from critical defects. First, no showing was made to the trial judge that the natural father cannot be located or is otherwise unavailable to fulfill the support obligation the law imposes on him. For purposes of this litigation at least, the identity of the biological father is known, and appellee cites no case law permitting the transfer of that obligation to one not standing in loco parentis based merely on the supposition— as stated by the trial judge — that “[i]f there ever was a possibility of locating and obtaining support from [the father], it is now surely negligible.” See Miller, 478 A.2d at 359 (permitting imposition of support duty based on financial detriment if parent “cannot locate the other natural parent” (emphasis added)).
Furthermore, the record fails to demonstrate that the child’s financial circumstances would be fundamentally altered by not imputing the support obligation to K.A.T. Other than as regards the two-year marital period before the parties’ separation, the mother *574presented no evidence — nor did the trial judge find — that she had received regular contributions from K.A.T. towards sharing a household or rearing the child. See note 2, supra. At most he made gifts to the child of clothing, toys, and money for recreational activities, and promised at the time of marriage to assist with her eventual college expenses. Whether or not the latter would support a claim of promissory estoppel with respect to college tuition, for example, an issue we do not decide,7 it is insufficient either alone or in combination with K.A.T.’s other contributions to establish the required detriment that would flow from failure to impose a general, continuing duty of child support on K.A.T.
Concluding as a matter of law that K.A.T. was not equitably estopped from denying paternity, we reverse the contrary judgment of the Superior Court and remand for further proceedings consistent with this opinion.
So ordered.
. Although, as is usual in such circumstances, we defer to the trial court's factual findings underlying its determination, see, e.g., L.C.D. v. District of Columbia, 488 A.2d 918, 922 (D.C.1985), our discussion in Dews, supra, makes clear that we review de novo the ultimate conclusion that equitable estoppel is or is not applicable on the given facts.
. At about the time of the child's birth, the mother applied for public financial assistance, listing D.P. as the child’s father. The mother apparently worked until 1978 when she became afflicted with a disease; thereafter she supported herself and the child through a combination of disability insurance, public aid for the dependent child, and food stamps.
. K.A.T. also claimed the child as a dependent on his medical insurance.
. Appellee relies on our seemingly more receptive allusion to equitable estoppel in the paternity context in S.A. v. M.A., 531 A.2d 1246, 1250 (D.C.1987), but as we pointed out in Dews, supra, the discussion in S.A. was dictum and the court there "expressly refused to rely on that doctrine to support its ruling.'’ Dews, 632 A.2d at 1168 (quoting S.A., 531 A.2d at 1255).
. In Fuller the period of "holding out” the child as the putative father’s child was considerably shorter than in this case.
. This is not a case such as Pietros, supra, in which the trial judge found that the mother’s "decision to bear the child was a direct result of the [putative father’s] promises” to support the child. Pietros, 638 A.2d at 547.
. The trial judge did not reach appellee's separate claim of promissory estoppel in regard to the promise of college aid.