Background: The long-term observation of vegetarians in affluent countries can provide insight into the relative effects of a vegetarian diet and lifestyle factors on mortality. Methods: A cohort study of vegetarians and healthconscious persons in Germany was followed-up prospectively for 21 years, including 1,225 vegetarians and 679 health-conscious nonvegetarians. Standardized mortality ratios compared with the German general population were calculated for all causes and specific causes. Within the cohort, Poisson regression modeling was used to investigate the joint effects of several risk factors on overall and cause-specific mortality. Results: Standardized mortality ratios for all-cause mortality was significantly below 100: 59 [95% confidence interval (95% CI), 54-64], predominantly due to a deficit of deaths from circulatory diseases. Within the cohort, vegetarian compared with nonvegetarian diet had no effect on overall