MEMORANDUM

Alberto Cadena appeals the district court’s affirmance of the Commissioner of Social Security’s denial of his application for Disability Insurance Benefits and Supplemental Security Income payments under Titles II and XVI of the Social Security Act. We have jurisdiction under 28 U.S.C. § 1291 and we affirm.

Cadena first argues that the ALJ improperly rejected the opinion of Dr. Bert, one of Cadena’s treating physicians. Dr. Bert’s residual functional capacity (RFC) questionnaire contained limitations on walking, sitting, standing, and lifting, which, when combined with Cadena’s need to miss work and take unscheduled breaks, would prevent Cadena from engaging in full-time work. The ALJ accepted Dr. Nolan’s less-restrictive medical testimony, instead of Dr. Bert’s more-restrictive medical opinion. This resulted in an RFC without all of the limitations to which Dr. Bert had opined. This decision of the ALJ was supported by substantial evidence. See Valentine v. Comm’r of Soc. Sec., 574 F.3d 685, 690 (9th Cir.2009). Dr. Bert’s medical opinion was inconsistent with Cadena’s reported level of activity and evidence of improvement in Cadena’s more-recent treatment notes, as well as with Dr. Nolan’s opinion. The ALJ also found it significant that Dr. Bert’s surgery recommendation conflicted with another examining physician’s recommendation against surgery in favor of nonoperative therapies such as weight loss, physical therapy, and exercise. The ALJ did not err in rejecting Dr. Bert’s medical opinion.

Cadena’s second claim of error is that the ALJ improperly found Cadena not fully credible about the intensity and dura-

tion of his back and leg pain. The ALJ, however, gave several reasons for finding Cadena not entirely credible. First, Cade-na’s reported daily activities included cooking, cleaning, and gardening, which were inconsistent with pain at a level of intensity that would be disabling. Second, Cade-na’s treatment notes, which included signs of improvement and a lack of objective medical evidence showing disabling pain, undermined Cadena’s credibility. Finally, Dr. Nolan observed that Cadena exaggerated his symptoms by voluntarily twitching his leg and describing his pain as an “11” on a scale of 1 to 10. See Tonapetyan v. Halter, 242 F.3d 1144, 1148 (9th Cir.2001) (a “tendency to exaggerate” is a proper credibility consideration). In our view, the reasons cited by the ALJ constitute “specific, clear and convincing reasons” for declining to fully credit Cadena’s subjective pain testimony and, as a result, the ALJ did not err by rejecting Cadena’s testimony in part. See Vasquez v. Astrue, 572 F.3d 586, 592 (9th Cir.2009).

Cadena’s third claim is that the ALJ improperly rejected the lay witness statements submitted by Wendy Deriso and Dorothy Stevens. The ALJ considered the statements, but determined that they described Cadena as “considerably more limited than medical findings would support.” The ALJ cited the medical evidence, including treatment notes that described Cadena’s daily activities, and determined that the lay witness statements conflicted with the medical evidence. See Lewis v. Apfel, 236 F.3d 503, 511 (9th Cir.2001) (“One reason for which an ALJ may discount lay testimony is that it conflicts with medical evidence.”). The reason given by the ALJ was “germane” to both Deriso’s and Steven’s statements, see Bruce v. Astrue, 557 F.3d 1113, 1115 (9th Cir.2009), so the ALJ did not err in declining to fully credit the statements.

We agree with Cadena’s fourth claim that the ALJ posed an improper hypothetical to the vocational expert. Dr. Nolan’s medical views, to which the ALJ gave controlling weight, suggested that Cadena had limitations on pushing and pulling with his lower extremities. Nonetheless, the ALJ did not include the push/ pull limitation in the hypothetical to the vocational expert and the expert testified Cadena could work as a delivery driver. The vocational expert’s testimony on cross examination showed that Cadena likely could not perform delivery-driver work if he had a lower-extremity limitation, so the ALJ’s step four ruling did not reflect all of Cadena’s limitations and thus was not supported by substantial evidence. See Osenbrock v. Apfel, 240 F.3d 1157, 1164-65 (9th Cir.2001).

However, the ALJ’s alternative ruling at step five—that Cadena could perform light, unskilled work that existed in significant numbers in the national economy—renders the step four error harmless. See Tommasetti v. Astrue, 533 F.3d 1035, 1042 (9th Cir.2008) (“Although the ALJ’s step four determination constitutes error, it is harmless error in light of the ALJ’s alternative finding at step five.”). Cadena argues that the ALJ erred in finding at step five that Cadena could perform the position of merchandise marker, because that position involves the use of fine-motor skills such as frequent fingering. But none of the three physicians that completed an evaluation pertinent to determining the RFC for Cadena concluded that he suffered limitations on the use of his hands. Moreover, Cadena testified at the administrative hearing that he had no difficulty reaching, handling, or grasping things. It was not error for the ALJ to find Cadena capable of working as a merchandise marker because substantial evi- ■ dence supported the premise that Cadena was able to perform fine-motor tasks with his hands.

AFFIRMED.

This disposition is not appropriate for publication and is not precedent except as provided by Ninth Circuit Rule 36-3. . The parties are familiar with the factual and procedural history of this case and we do not recount that history in detail here. . We express no opinion on the continuing evidentiary value of so-called "Waddell’s tests.” Although the AU discussed Dr. Nolan’s administration of Waddell’s axial load test in his decision, the ALJ relied on substantial evidence apart from the axial load test in finding Cadena not credible.