A historic cohort study was performed to determine the prevalence and incidence rates of nerve function impairment (NFI) as demonstrated by sensory testing with a nylon monofilament and standard tests of motor function. The records of 396 new leprosy patients registering at Green Pastures Hospital, Pokhara, between January 1988 and January 1992 were analysed. The mean follow-up period was 21 months. In all, 36% (141/396) of patients had either sensory or motor function impairment at their initial examination. For each nerve the prevalence of sensory and motor impairment is reported separately. The posterior tibial nerve was the most frequently affected (sensory) nerve (21%). Sensory impairment of the ulnar nerve was found in 17% of the patients; 8.8% had sensory impairment of the median nerve. The overall incidence rate of motor function impairment was 7.5 (5.4-10) per 100 person years at risk (PYAR). Sensory impairment had a significantly higher rate of 13 (10-17)/100 PYAR (rate ratio (1.8 (1.2-2.7), p = 0.0076). Bl patients had a significantly higher incidence rate of nerve function impairment than BT patients (rate ratio 2.3 (1.4-3.7), p = 0.006). Altogether 152/396 (39%) of the patients required corticosteroid treatment for 'recent' or 'acquired' impairment, and 78 of the patients (20%) developed severe nerve function impairment during or after antileprosy treatment. Analysis of potential risk factors for nerve function impairment showed a significant association with the extent of clinical disease expressed as the number of body areas (out of 9) with primary or secondary signs of leprosy (rate ratio 5.0 (1.5-17), p = 0.0091). It was concluded that nerve function impairment is a serious problem, often occurring during or after multidrug therapy. The extent of clinical disease expressed as a count of body areas involved, or of skin or nerve lesions may identify patients who are at increased risk of nerve damage.