To determine prevalence and correlates of US medical students’ self-identification as vegetarians.

Medical students were anonymously surveyed via questionnaire three times: at freshmen orientation, orientation to wards, and during senior year.

Medical students in the Class of 2003 (n=1,849) at 15 US medical schools (response rate 80%).

We examined self-reported vegetarianism, abstinence from meat items on a food frequency questionnaire, and associations between students’ vegetarianism and their health-related outcomes.

Bivariate associations were tested with χ² tests.

During medical school, 7.2% of students self-identified as vegetarians; this percentage declined over time. Those who were vegetarians for health reasons (66% of vegetarians) ate more fruits and vegetables than those who were vegetarians for nonhealth reasons (P=0.02). Vegetarians were more likely (P<0.01) than nonvegetarians to eat more fruits and vegetables (P=0.002); be women (P=0.009); be Hindu, Buddhist, or Seventh Day Adventist (P≤0.0004); be politically liberal (P=0.007); have a body mass index ≤25 (P=0.008); or, as freshmen, to perceive nutrition counseling as highly relevant to their intended practices (P=0.007). Vegetarian students were no more likely to counsel patients about nutrition than were nonvegetarians.

Prevalence of vegetarianism was higher among US medical students than among other US adults, although the prevalence declined during medical school. Medical students and physicians with healthful personal practices are more likely to encourage such behaviors in their patients, although the specific nutrition habit of vegetarianism among medical students was unassociated with their nutrition counseling practices.