AIMS: To determine if inequality in height between claws within each hindlimb of dairy cattle is a risk factor for the development of lameness and to determine the effect of preventive trimming of these claws on the subsequent risk of lameness. METHODS: Cows (n=2,695) on three dairy farms in Canterbury, New Zealand, were examined and the height differential between their hind claws on each limb was measured. Cows were blocked by age, then breed and grouped by height differential, before being randomly assigned to either treatment or control groups. Cows in the control group were not trimmed, whilst cows in the treatment group were all trimmed according to the 'Dutch Trimming' method by an experienced foot trimmer. All subsequent clinical incidences of lameness were identified, recorded and classified according to five categories. The overall incidence of lameness, the effect of claw height differential on the incidence of lameness and the effect of trimming on the incidence of lameness were assessed using both GEE modelling and survival analysis. RESULTS: The incidence of lameness identified in any untrimmed hindlimb irrespective of diagnosis, measured at a limb level and assessing the first incidence of lameness only, was 10.0% for a 6-month period from November to drying off. The predominant lesion identified was white line disease (WLD; 7.9% incidence and 78.6% of all lameness). The overall incidence of hindlimb lameness measured at cow level for the same period was 14.5%. The risk of lameness was influenced by breed (p=0.016), age (p=0.002) and claw height differential (p=0.026). There were no significant interactions. The risk of lameness was lowest where the claw height differential was 0-2 mm. The incidence of lameness in the first 70 days following trimming was higher in older cows than in younger cows (p<0.001). Trimming did not affect lameness incidence (p=0.185). The Cox proportional hazard model demonstrated that farm (p<0.001), trimming group (p=0.021); and age (p=0.021) significantly affected the median days to lameness in the first 70 days following trimming. CONCLUSIONS AND CLINICAL RELEVANCE: Hind feet with lower height differentials between claws had a lower incidence of lameness. Trimming of these feet to remove the height differential did not reduce their overall incidence of lameness, but did increase days to first lameness.