In a prospective study (1961-1972) the incidence of rheumatic fever (RF), development of rheumatic heart disease (RHD) and penicillin prophylaxis efficacy were investigated in 300 000 children and 70 000 adults. Only patients whose index attack met the modified Jones' criteria were included; all were enrolled in a prophylaxis programme (i. m. benzathine-penicillin). The RF incidence decreased from an initial 61.2 attacks per 100 000 in children and 42.9 in adults to below 10 by 1969 and remained low thereafter. In total, 541 children (209 with carditis) and 176 adults (108 with carditis) were followed up for 5.5 years (average) after an index attack. At final examination, no valvular involvement was detected where previous attacks had been without clinical carditis. Patients studied from time of the primary attack with carditis (211) had a very favorable prognosis: only up to 3.5% had developed a moderate to severe heart lesion by the end of the study. However, of 68 adult patients followed up only from the time of recurrence with carditis and often having had a prolonged rheumatic history: 23.5% had a moderate to severe heart lesion and a further 20.6% had died of RHD by the end of the study. Prophylaxis efficacy calculated as a cumulative attack-recurrence rate per 100 patients and over a 5-year interval was 2.72 in protected as against 64.75 in unprotected children (mostly failing to comply with prophylaxis); in adults, the respective values were 1.3 and 24.92. The prevalence of serious residual RHD may be expected to decrease substantially in the future, starting with the younger age groups.