To examine prospectively the relationship between vegetable consumption and long-term survival. In 1965, a total of 1536 Italian males from two Italian rural cohorts of the Seven Countries Study, aged 45-65 years, were examined. Information on lifestyle and food consumption collected at this visit, and total and cause-specific mortality data collected in 30 years of follow-up were analyzed for the present study. During a period of 30 years, 1096 deaths occurred (308 from coronary heart disease, 325 from cancer, 158 from cerebrovascular disease and 305 from all other causes). The age-adjusted life expectancy for men consuming more than 60 g/day of vegetables was nearly 2 years longer than for men consuming less than 20 g/day. This increase in survival was more striking in smokers than nonsmokers (2.1 vs. a 1 year gain). The association also held for both geographic cohorts, although the pattern of vegetable consumption was very different in the two villages. The results suggest a positive association between vegetable intake and life expectancy. Vegetable intake may be especially protective for smokers although the biological explanation for such an effect is unclear.