The objective of this study was to compare the time it took for sows to approach feed during different lameness states. Twenty-one, clinically healthy, mixed-parity, crossbred sows (220.4 ± 21.3 kg) were used. The sow was the experimental unit and a cross-over design with 3 (days) x 2 (hooves) factorial arrangement of treatments was used. Each sow served as her own control and treatment. On induction day, 10 mg of amphotericin B was injected in the distal interphalangeal joint space in both claws of one hind hoof. The sows were limited fed and latency to approach the first feeding in the home pen after feed presentation was measured on D-1 (1 day before induction), D+1 (1 day after injection of amphotericin B to induce lameness), and D+6 days (6 days after the induction of lameness). Video recordings were continuously collected on the three days from the hours 0600 to 1800 and later analyzed by a single trained observer. Data were analyzed using the PROC MIXED procedure in SAS. A P value of ≤ 0.05 was considered to be significant. A total of 14% of sows were classified as did not approach on the D+1. Not including sows that did not approach, sows took approximately 5 seconds longer to approach feed on D+1 compared to D-1 (P = 0.0013), but latency to approach feed did not differ between D-1 and D+6 (P = 0.80) In conclusion, when sows were lame they took longer to approach feed relative to D-1 and D+6. With this difference being small, time to approach feed may not be a sensitive behavioral bench mark for producers to use on farm to distinguish between sound and lame sows.