A nested case-control study of 84 incident cases of patients with idiopathic Parkinson's disease (PD) detected by June 30, 1994 and 336 age-matched control subjects, compared previously-documented intake of total dietary vitamin E and of selected vitamin E-containing foods. All study subjects had been followed for 27 to 30 years after diet recording in the 8,006-man Honolulu Heart Study cohort. We determined PD outcomes by periodic cohort re-examination and neurologic testing, private physician reports, examination of O'ahu neurologists' office records, and continual death certificate and hospital discharge diagnosis surveillance. Data on vitamin E intake, obtained from three dietary data sets at the time of cohort enrollment (1965 to 1968), included a food-frequency questionnaire and a 24-hour photograph-assisted dietary recall administered by trained dietitians. Although absence of PD was significantly associated with prior consumption of legumes (adjusted OR = 0.27, 95% CI 0.09 to 0.78), a dietary variable preselected for high vitamin E content, neither food categories nor quartiles nor continuous variables of vitamin E consumption were significantly associated with PD occurrence. Though consistent with prior reports of PD protection afforded by legumes, and with speculation on the possible benefits of dietary or supplemental vitamin E in preventing PD, these preliminary data do not conclusively document a beneficial effect of dietary vitamin E on PD occurrence.