The purpose of this study was to assess the effect of lameness on dairy cow survival. Cox's proportional hazards regression models were fitted to single-lactation data from 2520 cows in 2 New York State dairy herds. Models were controlled for the time-independent effects of parity, projected milk yield, and calving season, and for the time-dependent effects of lameness and culling. Other common diseases were found to be nonconfounding and so were not included in any of the final models. Survival was measured as the time from calving until death or sale. Cows were censored if they reached the start of the next lactation or end of the study, whichever occurred first. All models were stratified by herd. For all lameness diagnoses combined, survival in the herd decreased for those cows becoming lame during the first half of lactation, with a hazard ratio of up to 2 times that of a nonlame cow. Foot rot diagnosed during the second or third months of lactation decreased survival during the same time period (hazard ratio=5.1; 95% confidence interval=1.6 to 16.2). Sole ulcers diagnosed in the first 4 mo of lactation decreased survival in several subsequent periods in which the strongest association was between diagnosis in the third and fourth months of lactation and exit from the herd during that same period (hazard ratio=2.7; 95% confidence interval=1.3 to 6.0). Foot warts were not associated with decreased survival in this analysis. Lameness was never associated with increased survival in any of the models.