AIM: To compare different method(s) to detect peripheral neuropathy in leprosy and to study the validity of the monofilament test (MF) and the voluntary muscle test (VMT) as standard tests of nerve function. DESIGN: A multi-centre cohort study of 303 multibacillary (MB) leprosy patients. METHODS: Newly registered MB patients requiring a full course of MDT were recruited in two leprosy outpatient clinics in North India. Controls were people without leprosy or neurological conditions, attending the dermatological outpatient departments of the same clinics. Nerve function was evaluated electrophysiologically using standard parameters for sensory and motor nerve conduction (NC) testing, warm and cold detection thresholds (W/CDT), vibration perception thresholds, dynamometry, MF and VMT. The latter two defined the outcomes of sensory and motor impairment. RESULTS: 115 patients had nerve damage or a reaction of recent onset at diagnosis. Sensory and motor amplitudes and WDTs were the most frequently abnormal. Among the nerves tested, the sural and posterior tibial were the most frequently impaired. In the ulnar nerve, sensory latencies were abnormal in 25% of subjects; amplitudes in 40%. Ulnar above-elbow motor conduction velocities were abnormal in 39% and amplitudes 32%. WDTs were much more frequently affected than CDTs in all nerves tested. The thresholds of all test parameters differed significantly between controls and patients, while only some differed between patients with and without reaction. Good concordance was observed between MF results and sensory latencies and velocities (direct concordance 80% for the ulnar). However, a proportion of nerves with abnormal MF results tested normal on one or more of the other tests or vice versa. Concordance between VMT and motor conduction velocities was good for the ulnar nerve, but for the median and peroneal nerves, the proportion impaired by VMT out of those with abnormal motor conduction was very low. CONCLUSIONS: Concordance between monofilaments and other sensory function test results was good, supporting the validity of the monofilaments as standard screening test of sensory function. Concordance between VMT results and motor nerve conduction was good for the ulnar nerve, but very few median and peroneal nerves with abnormal conduction had an abnormal VMT. A more sensitive manual motor test may be needed for these nerves. Of the nerve assessment tests conducted, NC amplitudes and warm sensation were the most frequently affected. Therefore, nerve conduction studies and WDT measurements appear to be most promising tests for early detection of leprous neuropathy. The pattern of concordance between tactile and thermal sensory impairment failed to support the hypothesis that small fibre neuropathy always precedes large fibre damage. Warm sensation was more frequently affected than cold sensation. This could indicate that unmyelinated C fibres are more frequently affected than small myelinated Asigma fibres.