dissenting:
I respectfully dissent because, in my view, substantial evidence supports the decision of the Administrative Law Judge (ALJ). Unlike the majority, I am persuaded that the ALJ gave proper weight to the opinions of all medical providers.
For purpose of this appeal, Social Security claimant Karen Ryan (Ryan) began visiting her treating physician, Dr. Moni-gatti-Lake regarding “her situation at work” on October 18, 1999. Ryan informed Dr. Monigatti-Lake that Ryan was on administrative leave following a random drug test during which she tested “positive for THC and apparently some amphetamines.” Dr. Monigatti-Lake diagnosed Ryan as experiencing a “stressful situation due to Ryan’s work difficulties,” which occurred “a couple of months” before “her 5 yr. retirement contract.”
Although Ryan expressed optimism about continuing with her employment, it was not to be. Ryan was terminated, and visited Dr. Monigatti-Lake on December 14, 1999, complaining of feelings of immobility, panic attacks, [and] crying spells. Ryan reported that Effexor1 she was taking was not helping. Dr. Monigatti-Lake observed that Ryan was “coherent and oriented,” assessed that Ryan was suffering from anxiety disorder, increased the Effexor prescription and prescribed Xa-nax.2
Ryan was seen a week later, at which time she reported that she could not concentrate very well and “would still get somewhat emotional,” but was doing better. Dr. Monigatti-Lake observed that Ryan was “calm” but “somewhat anxious.” Dr. Monigatti-Lake’s diagnosis was “[a]nx-iety disorder, improving” and “[depression, improving.” (Emphasis added). Dr. Monigatti-Lake encouraged Ryan to seek counseling, and gave her the names of some therapists.
*1203Ryan next visited Dr. Monigatti-Lake two and one-half months later, reporting that she had “bad days and good days.” Ryan inquired about BuSpar,3 which Dr. Monigatti-Lake thought was a good idea. Dr. Monigatti-Lake observed that Ryan was coherent “but very agitated” and diagnosed her with “[a]nxiety disorder.” Dr. Monigatti-Lake continued Ryan on Effe-xor, started a prescription for BuSpar and instructed Ryan that she could continue to take Xanax.
Two weeks later, Ryan was seen again. Ryan reported that she was doing well on BuSpar and was taking Effexor regularly. Dr. Monigatti-Lake observed that Ryan was coherent but very agitated. Ryan was diagnosed as suffering from anxiety disorder and again referred for counseling.
Approximately three weeks later, Dr. Monigatti-Lake assessed Ryan with anxiety disorder. Ryan reported that she was continuing to take her medications, had not yet contacted a counselor, was going out for two to three hours and was working in the garden. Although Dr. Monigat-ti-Lake referred Ryan to the Department of Social Services approximately one month later, she did not opine that Ryan was disabled. Approximately six weeks later, during a follow-up visit for knee surgery, Dr. Monigatti-Lake noted that Ryan had “quite a bit of anxiety” but was tolerating it “fairly well.” Dr. Monigatti-Lake diagnosed “anxiety disorder, improving.” (Emphasis added).
Over seven months later, Dr. Monigatti-Lake saw Ryan, observing that Ryan was calmer than normal. Dr. Monigatti-Lake continued to diagnose anxiety disorder. Three weeks later, Dr. Monigatti-Lake noted that Ryan was “slightly less anxious.” Dr. Monigatti-Lake diagnosed “anxiety disorder, slowly improving.” (Emphasis added). Over seven months later, Dr. Monigatti-Lake diagnosed Ryan with chronic depression and anxiety. A visit four months later reflected no indication of anxiety or depression.
Dr. Monigatti-Lake’s final notation approximately five months later stated that Ryan was still suffering from “occasional panic attacks.” At that time, Ryan reported that she was “still trying to get disability benefits.”
In summary, from December 14, 1999, through June 13, 2002, Dr. Monigatti-Lake diagnosed Ryan with anxiety, depression and panic attacks and noted continuous improvement. Not once during that period did Dr. Monigatti-Lake opine that Ryan was disabled, unable to work or agoraphobic.
In May, 2000, during the time she was being treated by Dr. Monigatti-Lake, Ryan underwent a psychiatric evaluation by Dr. Rajinder Randhawa. Dr. Randha-wa concluded that Ryan “does have the ability to perform and understand simple, repetitive, detailed and complex tasks which she is likely to use [sic] interest in due to lack of motivation and initiative and limited desires at present due to continuing depression and extreme anxiety at times.” (Emphasis added). Dr. Randha-wa’s assessment that Ryan would be unable “to maintain regular attendance in the workplace” was limited to that particular moment in time. In fact, he predicted that “[i]f she gets psychotherapy along with aggressive psychiatric treatment in terms of anti-anxiety and an anti-depressant medication in appropriate and adequate dosages[,] ... [h]er condition is likely to improve” and she could return to work.
*1204Drs. Harman and Harrison, consulting physicians, made the following functional assessment between June and October, 2000: “with continued [treatment] anticipate able to do complex in low public [sic] with moderate impairments in ability to sustain [concentration and attention] [persistence and pace] and ability to complete regular work-week without interruption from her [psychological symptoms].” This assessment was entirely consistent with Dr. Randhawa’s opinion that although Ryan was presently “unable to maintain regular attendance in the workplace[,] [h]er condition was likely to improve.” (Emphasis added).
On August 4, 2000 (between the times of Dr. Randhawa’s assessment and the completion of the Harman-Harrison assessment), Ryan was seen for a lesion on her leg. During that visit, she did not report any anxiety-related symptoms.
In December, 2000, Ryan was examined by Dr. Espino. Dr. Espino noted a history of anxiety disorder, and completed a functional capacity assessment. Although he listed several restrictions with regard to physical activity, his only reference to Ryan’s- psychological impairments was to state that she “clearly needs to be on her medication for anxiety.”
On December 24, 2002, Ryan was seen by Colleen McKinnon, a licensed social worker with the Nevada County Behavioral Health Department, who noted a history of anxiety disorder and major depression.
Ryan’s final examination was done by Dr. Crisp, a physician with the Nevada County Behavioral Health Department. Ryan reported that she could relax at home but got anxious when she went out. Taking Ryan’s report at face value, Dr. Crisp diagnosed her with major depression with agoraphobia and anxiety.
Ignoring the lack of a finding of disability by any of the medical providers, the majority opinion discounts the substantial evidence supporting the Administrative Law Judge’s (ALJ) denial of benefits. The majority opinion rests on the ALJ’s rejection of Dr. Randhawa’s observation that Ryan “would have difficulty maintaining regular attendance in the workplace due to anxiety and depression and due to interruptions from her psychiatric eondi-tions[,]” and on the ALJ’s discounting of Dr. Crisp’s diagnosis of agoraphobia. See Majority Opinion at pages 1198-1202.
In reviewing this matter, it is important to clarify the standard of review. In Social Security cases, we employ a hierarchy of deference to medical opinions depending on the nature of the services provided. We “distinguish among the opinions of three types of physicians: (1) those who treat the claimant (treating physicians); those who examine but do not treat the claimant (examining physicians); and those who neither examine nor treat the claimant (nonexamining physicians).” Lester v. Chater, 81 F.3d 821, 830 (9th Cir.1996) (footnote reference omitted). A treating physician’s opinion is entitled to more weight than an examining physician’s opinion, and an examining physician’s opinion is entitled to more weight than a non-examining physician’s opinion. See id. As applied to this ease, Dr. Monigatti-Lake’s opinion was entitled to the most weight due to her status as treating physician. Acceptance of Dr. Monigatti-Lake’s opinion over that of Drs. Randhawa and Crisp, examining physicians, would be entirely consistent with our governing precedent. See Thomas v. Barnhart, 278 F.3d 947, 957 (9th Cir.2002) (noting the deference given to a treating physician’s opinion). Indeed, crediting the opinions of Drs. Randhawa and Crisp over that of Dr. Monigatti-Lake absent “specific and legitimate reasons supported by substantial evidence in the record” would directly contravene our precedent. Lingenfelter v. Astrue, 504 F.3d 1028, 1038 n. 10 (9th Cir.2007) (cita*1205tions omitted). Yet that is precisely what a majority of the panel concludes that the ALJ should have done and does itself. I respectfully disagree because the record does not support crediting the opinion of those examining doctors over the treating doctor.
An examination of the record reveals that Dr. Monigatti-Lake, Ryan’s treating physician, never opined that Ryan was unable to work or would “have difficulty maintaining regular attendance in the workplace” as a result of her improving anxiety disorder and treatable depression. In fact, Dr. Monigatti-Lake’s final notation regarding Ryan reflected only “occasional panic attacks” and an observation that Ryan was “still trying to get disability benefits.” Although this notation reflects Dr. Monigatti-Lake’s awareness of the possibility of a disability diagnosis, she did not in any way indicate that Ryan’s conditions were disabling, although she had treated those conditions for a period in excess of two and one-half years.
A comparison of the medical records of the treating physicians from December 14, 1999, through June 13, 2002, with the records of Dr. Randhawa, a one-time examining physician, reveals a conflict to the extent of Dr. Randhawa’s opinion that Ryan would “have difficulty maintaining regular attendance in the workplace.” Indeed, Dr. Monigatti-Lake’s notes indicate the opposite expectation. Dr. Monigatti-Lake expressly noted on several visits that Ryan’s condition was improving, with her final notation referencing only “occasional panic attacks.”
Given the conflict in the record between the treating physician’s observation of a non-disabling, improving condition and the examining physician’s opinion that Ryan’s condition disabled her from regular attendance at work, it fell to the ALJ to resolve the conflict. See Edlund v. Massanari, 253 F.3d 1152, 1156 (9th Cir.2001). Resolving the conflict in favor of the treating physician’s observations was eminently proper. Indeed, “[wjhere the evidence can reasonably support either affirming or reversing the decision, [this Court] may not substitute [its] judgment for that of the Commissioner.” Parra v. Astrue, 481 F.3d 742, 746 (9th Cir.2007) (citation omitted). Yet that is precisely what the majority does by selectively crediting the opinion of examining doctors over that of a long-term treating doctor. The majority opinion states that nothing in the record indicates that Dr. Monigatti-Lake expressed an opinion that Ryan was capable of maintaining a regular work schedule. See Majority Opinion, p. 1201 n. 6. However, Dr. Monigatti-Lake’s notations of continuing improvement are the very antitheses of a disability. See, e.g. Rollins v. Massanari, 261 F.3d 853, 856 (9th Cir.2001) (recognizing the inherent conflict between an observation of improvement and a finding of disability); see also Johnson v. Shalala, 60 F.3d 1428, 1433 (9th Cir.1995) (same).
The majority opinion takes issue with the ALJ’s decision to give less weight to Dr. Randhawa’s assessment because it was “based more upon the claimant’s subjective complaints which are not fully supported in the record.” However, this conclusion by the ALJ is amply supported by the record. Dr. Randhawa’s conclusion that Ryan could not maintain regular attendance was explicitly predicated on the following statements: “[Ryan] said it is very difficult for her to even grow up [sic] an application as she starts having severe palpitations. Cold sweat and extreme anxiety and now she feels like she is going to pass out.” (Emphases added). Dr. Randhawa did not connect his conclusion to any objective medical findings or observations. Rather, as the ALJ correctly noted, these quintessentially subjective complaints are not a sufficient basis of support for a dis*1206ability determination. One need look no further than the case cited by the majority, Bayliss v. Barnhart, 427 F.3d 1211 (9th Cir.2005) for guidance. In Bayliss we specifically ruled that substantial evidence supports the ALJ’s decision not to rely on a medical opinion based on subjective complaints rather than clinical evidence. See id. at 1217. Like the doctor in Bayliss, Dr. Randhawa did not link his opinion to any clinical findings. Rather, he referenced only Ryan’s subjective complaints. In such a circumstance, the ALJ’s decision to give less weight to Dr. Randhawa’s medical opinion is in accord with our precedent. See id.
The majority opinion also criticizes the ALJ’s reliance on Dr. Monigatti-Lake’s treatment records. In the majority’s view, a diagnosis of “anxiety disorder” and/or “depression” by Dr. Monigatti-Lake obviates any conflict between Dr. Randhawa’s opinion that Ryan would be unable to maintain regular attendance in the workplace and Dr. Monigatti-Lake’s observations of improving functional capacity. However, a diagnosis of anxiety disorder or depression is a far cry from a finding of disability. In fact, millions of people work every day while being treated for anxiety disorder and/or depression. See, e.g., Lynn Elinson, Patricia Houck, Steven C. Marcus and Harold Alan Pincus, Depression and the Ability to Work, PsyCHiatric SeRV. 55, 29-34 (2004), available at http:// www.ps.psychiatryonline.org/cgi/content/ full/55/1/29 (“Depression treatment has ... been shown to be cost-effective, to keep depressed persons employed, and to improve the productivity of depressed persons who are already working.”) (footnote references omitted).
More to the point, the reason a treating physician’s opinion is weighed more heavily is because the treating physician has a sustained relationship with the patient and is more likely to have a complete and accurate grasp of the patient’s medical condition. See Smolen v. Chater, 80 F.3d 1273, 1285 (9th Cir.1996); see also 20 C.F.R. § 404.1527(d)(2). Yet the majority discounts the long-term medical record developed by the treating physician Dr. Mon-igatti-Lake in favor of a conflicting medical opinion predicated on a single visit. This conclusion turns our Social Security jurisprudence on its head. Conflict arose when the examining physician, Dr. Ran-dhawa, used Ryan’s subjective complaints to opine that she was unable to maintain regular attendance at work even though the treating physician consistently noted that Ryan was coping fairly well with her conditions. It was entirely within the ALJ’s purview to resolve the conflict. See Edlund, 253 F.3d at 1156.
In an effort to bolster its unjustifiable reliance on the onetime examination of Dr. Crisp to support a diagnosis of agoraphobia, the majority opinion cites Regennitter v. Commissioner, 166 F.3d 1294, 1299 (9th Cir.1999). However, that case is readily distinguishable. In Regennitter, the later physician added diagnoses of post-traumatic stress disorder and nightmare disorder to the related diagnoses of major depression and panic disorder made by both doctors. See id. In this case, both Drs. Monigatti-Lake and Randhawa diagnosed anxiety disorder and depression. Dr. Crisp was the only medical provider in the entire record who diagnosed agoraphobia. Contrary to the majority’s assertion, this diagnosis was not only at odds with that of the treating physician, it was at odds with the entire medical record, providing ample support for the ALJ’s decision to give it less weight. See Thomas, 278 F.3d at 957. In addition and as the ALJ noted, the diagnosis was predicated on Ryan’s statements that she “was extremely anxious ... when she leaves home” and that “she has trouble even going to the grocery store.” As we have repeatedly held, an ALJ may discount a medical opinion that relies on *1207subjective statements rather than clinical findings. See Tonapetyan v. Halter, 242 F.3d 1144, 1149 (9th Cir.2001).
The majority’s citation to Young v. Heckler, 803 F.2d 963, 968 (9th Cir.1986) is inapposite. Reliance on Young would be appropriate if the evidence in the record reflected that Ryan’s condition was “progressively deteriorating.” Id. However, the medical record in this case is to the contrary. The only physician who had a progressive perspective of Ryan’s condition was Dr. Monigatti-Lake, her treating physician. And Dr. Monigatti-Lake consistently noted that Ryan’s condition was improving, rather than deteriorating.
In sum, because Drs. Randhawa’s and Crisp’s opinions were based on Ryan’s subjective complaints and conflicted with that of the treating physician, substantial evidence supported the respective weights given to those opinions by the ALJ when he determined that Ryan was not entitled to Social Security benefits. Unlike the majority, I would adhere to our precedent, afford the treating physician’s opinion the greater weight to which it is due, and affirm the district court.
. Effexor (Venlafaxine) is used to treat, depression, anxiety disorder, and panic disorder. American Society Of Health-System Pharmacists, Inc., Drug Information: Venlafaxine (2007), http://www.nlm.nih.gov/ medline-phis/drugmfo/medmaster/a694020.html.
. Xanax (Alprazolam) is used to treat anxiety disorder and panic attacks. American Society Of Health-System Pharmacists, Inc., Drug Information: Alprazolam (2003), http://www.nlm. nih.gov/medlineplus/druginfo/medmaster/a 684001.html.
. BuSpar (Buspirone) is used to treat anxiety disorders or in the shortterm treatment of symptoms of anxiety. American Society Of Health-System Pharmacists, Inc., Drug Information: Buspirone (2003), http://www.nlm.nih. gov/medlineplus/druginfo/medmaster/a 688005.html.