A prospective study was conducted in Takayama, Gifu, Japan, to evaluate the association between diet and the onset of menopause. A total of 1,130 female residents aged 35-54 years who were premenopausal and completed a validated semiquantitative food frequency questionnaire in 1992 were contacted by mail with a follow-up questionnaire in 1998 to update information on menopause. Onset of menopause was defined as a woman's age at the last menstrual period prior to stopping menstruation for 12 months. During the 6-year study period, 296 women experienced natural menopause. The Cox proportional hazards model was used to estimate hazard ratios of the onset of menopause after controlling for age, total energy, body mass index, years of smoking, and age at which regular menstrual cycle began. The authors found that green and yellow vegetable intake was significantly inversely associated with the 6-year incidence of menopause (hazard ratio = 0.71, 95% confidence interval: 0.54, 0.95 for the highest vs. lowest tertile of intake, p for trend = 0.02). Association of carotene intake with the incidence of menopause was of borderline significance (hazard ratio = 0.78, 95% confidence interval: 0.59, 1.04, p for trend = 0.07).