dissenting:
Although this is a close case, given our deferential standard of review I cannot concur with the panel majority’s decision to reverse and to remand for further proceedings. We must affirm the ALJ’s decision if it (1) employs the correct legal standard, and (2) is supported by substantial evidence, which amounts to more than a scintilla, but less than a preponderance. See Ryan v. Comm’r of Soc. Sec., 528 F.3d 1194, 1198 (9th Cir.2008); Connett v. Barnhart, 340 F.3d 871, 873 (9th Cir.2003). Because the ALJ properly evaluated the testimony of Webber and her husband as well as the opinion of her treating physician, I respectfully dissent.
I
Webber was insured only through September 30, 2001. Thus, the question presented is whether substantial evidence supports the ALJ’s findings as to Webber’s disability on that date; her disability after that date is irrelevant. This distinction is overlooked in the panel majority’s conclusion that the ALJ failed to adduce “specific, clear and convincing reasons” for discrediting Webber’s testimony on the ba*315sis of her reported daily activities. Smolen v. Chater, 80 F.3d 1273, 1281 (9th Cir.1996).
A claimant’s daily activities are relevant to the ALJ’s assessment of her disability and may impact the claimant’s credibility. See Fair v. Bowen, 885 F.2d 597, 603 (9th Cir.1989). Here, the ALJ noted that Webber performs many household activities, albeit in a limited capacity, and concluded that “the mental and physical capabilities” Webber employs to perform these tasks “replicate those necessary for obtaining and maintaining employment.” 1
Significantly, Webber’s testimony occurred on July 27, 2003. The ALJ took this into account, explaining that “[w]hile claimant and her husband were generally credible in their description of a medical condition that currently imposes significant limitations on claimant’s ability to sustain exertional activities, it also appears that the condition has been progressively deteriorating, and the effects were not as severe prior to September 30, 2001.” (emphasis added). The ALJ also explained that Webber’s daily activities “now appear to be limited by [her fibromyalgia],” (emphasis added), but noted that Webber’s own statements in 1998 indicated that she was able to perform her daily chores “as long as she took her time.” Indeed, the ALJ lists various events from 1995 through 2002 which suggest the progressive onset of symptoms.2
Relying on this assessment of the record as a whole, the ALJ concluded that while Webber’s testimony may very well indicate she was disabled as of July 27, 2003, it was not credible evidence as to the state of her disability on September 30, 2001.3
Accordingly, substantial evidence supports the ALJ’s decision to discredit Webber’s testimony.4
II
The panel also concludes that the ALJ erred in discrediting the opinion of Webber’s treating physician, Dr. Kirsten. An ALJ may reject the opinion of a treating physician in favor of the opinion of an examining physician if the ALJ makes “findings setting forth specific, legitimate *316reasons for doing so that are based on substantial evidence in the record.” Magallanes v. Bowen, 881 F.2d 747, 751 (9th Cir.1989) (internal quotation marks and citation omitted). Here, the ALJ gave “less weight” to Dr. Kirsten’s opinion because: (1) there were large gaps between Webber’s visits to Dr. Kirsten; and (2) Dr. Kirsten’s opinion was inconsistent with other medical evidence in the record.
A
The panel majority erroneously rejects the ALJ’s reliance on the gaps between Webber’s visits for failure to examine the explanation provided by Dr. Kirsten. Maj. Disp. at 313-14. Dr. Kirsten justified the gaps by explaining that “there [were] no new medication/modalities to try, and [Webber was] fairly stable.” As discussed above, supra at 315 n. 2, while this might be a valid reason not to schedule appointments, the fact remains that the absence of appointments undermines Dr. Kirsten’s reliability as a witness to Webber’s condition on September 30, 2001. The ALJ did not reject Dr. Kirsten’s reason for not scheduling more appointments; he concluded that because such appointments were infrequent, Dr. Kirsten’s 2003 retrospective analysis was less reliable than other evidence in the record. Such fact was a specific and legitimate reason for according Dr. Kirsten’s opinion less weight.
B
The panel majority also erroneously discounts the “inconsistencies noted by the ALJ” as “unsupported by the record.” Maj. Disp. at 314. At least one inconsistency, however, is supported by the record: Dr. Kirsten’s 2003 opinion states that Webber suffers from “poor muscle strength,” whereas Dr. McCarthy’s 1998 exam indicates that Webber had normal muscle tone and motor strength in her left arm.5 While Dr. McCarthy’s exam was conducted three-and-a-half-years before September 30, 2001, Dr. Kirsten did not complete her report until one-and-a-half years after such date. Dr. Kirsten stated that Webber had been disabled since 1995. Dr. McCarthy stated the opposite in 1998. By highlighting this conflict, the ALJ articulated a specific and legitimate reason not to credit Dr. Kirsten’s assessment.
Thus, the ALJ identified two specific and legitimate reasons for discrediting Dr. Kirsten’s 2003 report. Any reliance on other, improper grounds was harmless error because the remaining two reasons are “legally valid” justifications for discrediting an examining physician’s testimony. Carmickle, 533 F.3d at 1162.
m
Based on the foregoing,6 I conclude that the ALJ’s decision was supported by substantial evidence. I respectfully dissent from the court’s opinion to the contrary.
. The ALJ went on to discuss how Webber’s daily activities were “transferable to the workplace,” Maj. Disp. at 3, explaining that such activities indicated that as of September 30, 2001, she was capable of "a substantial range of sedentary exertional activity” pursuant to 20 C.F.R. § 404.1567(a).
. The ALJ also noted the large gaps between Webber’s visits with doctors. For example, Webber did not visit a physician for a twenty-one-month period surrounding her date of last insured. While Dr. Kirsten attempted to explain the reason for the infrequent visits, such gaps in the record nonetheless create difficulties in assessing the nature of Webber’s condition on September 30, 2001. Because claimants bear the burden of proving disability at this stage of the analysis, Burch v. Barnhart, 400 F.3d 676, 679 (9th Cir.2005), the cost of this lack of evidence falls on Webber.
. The panel is right to rule that the ALJ erred in relying on Webber’s failure to follow or seek treatment. Nonetheless, such reliance was harmless error. See, e.g., Batson v. Comm’r of the Soc. Sec. Admin., 359 F.3d 1190, 1195-97 (9th Cir.2004). Here, the ALJ's determination — that Webber was less than fully credible because her testimony conflicts with her daily activities and the medical evidence — is “legally valid” despite the fact that the ALJ erred in not addressing her explanation for not following recommended treatment protocols. See Carmickle v. Comm’r, 533 F.3d 1155, 1162 (9th Cir.2008); see also Thomas v. Barnhart, 278 F.3d 947, 959-60 (9th Cir.2002) (claimant’s performance of daily activities relevant to credibility); id. (inconsistencies between claimant’s testimony and medical evidence also relevant to credibility).
. The ALJ applied the same reasoning in discrediting the testimony of Webber’s husband. That determination, therefore, is likewise valid.
. Dr. McCarthy's assessment was not limited to Webber's left arm. Dr. McCarthy, a neurologist, performed a complete physical and cranial nerve examination and concluded that Webber’s neurological evaluation was normal and that she was likely depressed. Although Dr. McCarthy specifically discussed the muscle tone in Webber’s left arm (because it was the only extremity she complained of), he found the arm "normal" — a finding in legitimate conflict with Dr. Kirsten’s assertion that Webber had poor muscle strength since 1995.
. While the panel majority does not reach the question, I would rule that because the AU found Webber was able to perform a "substantial range of sedentary work,” reliance on the on the medical-vocational rules (Grids) was appropriate.