Changes over time in the prevalence and intensity of Schistosoma mansoni infection were measured by serial quantitative stool examinations using the modified Ritchie concentration technique in a 9-year prospective study of an endemic Puerto Rican community with a population of about 1,000 persons. The complete interruption of transmission was achieved by snail control during the 2nd year of the study, in February 1973. Annual stool specimens were obtained from all willing community residents. In addition, from 27 of these residents 10 consecutive stools were collected in each of 3 years: 1973, 1976, 1977. After a chemotherapy campaign with oxamniquine in 1980, only eight of these 27 persons remained untreated. Ten consecutive stools were collected from seven of these untreated individuals in 1981. We calculated the rate of decline (beta) in geometric mean egg count in the cohort of 27 over 5 years and in the subcohort of seven over 9 years. Similarly, beta was calculated from the change in the single annual stool counts in 528 persons providing data for all of the first 6 years of the study. Estimates of the average life-span (-1/beta, in a model assuming constant rate of death) of the adult S. mansoni with 95% confidence intervals are for the cohort of 27, 5.5 years (4.0 to 9.1), for the cohort of seven, 37 years (8.0 to infinity), and and for the cohort of 528, 35 years (16 to infinity).(ABSTRACT TRUNCATED AT 250 WORDS)