An epidemiological survey has been conducted during a 3-month period in the Regional Hospital of S. Giovanni Rotondo, a small city in Southern Italy. A total of 1,878 patients aged 12 years and over admitted to the hospital were randomized during the study period. Each patient was asked 7 specific questions focusing on symptoms most likely due to polyneuropathy. The sensitivity and specificity of the screening procedure were tested separately and found to be high. All cases who gave affirmative answers to at least 2 questions were carefully investigated in search of a polyneuropathy. Of the 20 cases which gave an affirmative answer to 2 or more questions, 19 (1 % or 10 cases per 1,000 hospital population) had a neurological examination fitting with the diagnosis of polyneuropathy. There were 3 males and 16 females with a median age of 61 years. Diabetes was the commonest associated disorder. The disease ran a mild course in 12 cases and was severe in only 1. The screening model used in the current study seems a valuable and simple approach for case ascertainment in epidemiological investigations of polyneuropathy.