The epidemiology of Guillain-Barré syndrome was examined in a population-based retrospective study using defined diagnostic criteria and linked hospital records. Between 1974 and 1986 there were 72 incident cases, giving a crude annual incidence rate (95 per cent confidence interval) of 1.1 per 100,000 persons per year (0.8-1.4). Age-adjusted incidence rates were insignificantly higher for women, 1.23 (0.8-1.6), than for men, 1.0 (0.6-1.3). After infancy, rates increased with age, and remained high in the over-75 age group. These variations in incidence were not explained by variations in severity. There was evidence of an increase in rates over the period surveyed. Rates tended to be higher in the winter and early spring. The incidence of Guillain-Barré syndrome in the elderly has been underestimated in the past, and this study shows no evidence of a previously reported bimodal distribution of age-specific incidence in adult life.