An epidemiological study of parkinsonism over a 13-year period (1967 through 1979) is presented, updating previous reports on incidence and trend in the population of Rochester, Minnesota. The overall average annual incidence of parkinsonism per 100,000 population was 20.5, adjusted to the 1970 total United States population, which is virtually unchanged from previous observations. Incidences calculated for each calendar year (1967 through 1979) revealed no remarkable change following the 1976 swine flu vaccination program. There was no sex difference and the peak incidence occurred between ages 75 and 84 years. Idiopathic Parkinson's disease was the most common variant (86%), followed by drug-induced parkinsonism (7%). There were no new cases of postencephalitic parkinsonism diagnosed during the study period. Relative frequency of other types of Parkinson's disease as identified by practicing neurologists is presented. For each case two age- and sex-matched controls were selected from the Rochester population. The survival rates in the controls were comparable to the general population of the west north central region of the United States. The mortalities in the patients were significantly higher (p = 0.001) than the controls and were unchanged from previous rates reported from the same community. In the 69 (50%) patients treated with levodopa, the mortality was comparable to that in controls. The favorable outcome in these cases is attributed to bias resulting from selection of healthier patients for treatment.