The results of a population-based follow-up study are presented. The baseline study which started in August 1987, was carried out to determine the prevalence of human immunodeficiency virus (HIV-1) infection in the Kagera region of Tanzania. A multistage cluster sampling technique was adopted in the selection of the study population. In the follow-up survey which started in June 1988, members of the same study population were revisited and studied in order to determine the incidence of HIV-1 infection among those who were HIV seronegative in the initial survey. HIV serology was conducted by using enzyme-linked immunosorbent assay and all positive sera were confirmed by the Western blot technique. A total population of 1316 adults aged 15-54 years was studied, constituting an average follow-up response rate of 69% in the rural areas and of 59% in the urban area. The overall incidence of HIV-1 infection among the adult population sample was 13.7 per 1000 person-years at risk with the highest incidence in the urban zone (47.5 per 1000 person-years at risk) and the lowest incidence in one of the three rural zones (4.9 per 1000 person-years at risk). The age-specific annual incidence was highest in the age group 25-34 years for males and in the age group 15-24 for females. From these results it was estimated that the number of newly infected adults in the region each year is about 8200 with a range between 5400 and 11,000 corresponding to the 95% confidence limits on the overall incidence.(ABSTRACT TRUNCATED AT 250 WORDS)