To assess the reliability and validity of self-reported height and weight, and variables calculated from these values, in a diverse sample of adolescents. A convenience sample of students (n = 4619) in grades 9 through 12 reported their height and weight on two questionnaires administered approximately 2 weeks apart. Using a standard protocol, a subsample of these students (n = 2032) also were weighed and had their height measured following completion of the first questionnaire. Self-reported heights at Time 1 and Time 2 were highly correlated, and the mean difference between height at Time 1 and Time 2 was small. Results were similar for self-reported weight at Time 1 and Time 2 and body mass index (BMI) calculated from these values. Although self-reported values of height, weight, and BMI were highly correlated with their measured values, on average, students overreported their height by 2.7 inches and underreported their weight by 3.5 pounds. Resulting BMI values were an average of 2.6 kg/m(2) lower when based on self-reported vs. measured values. The percentages of students classified as "overweight" or "at risk for overweight" were therefore lower when based on self-reported rather than on measured values. White students were more likely than those in other race/ethnic groups to overreport their height, and the tendency to overreport height increased by grade. Female students were more likely than male students to underreport their weight. Self-reported height, weight, and BMI calculated from these values were highly reliable but were discrepant from measured height, weight, and BMIs calculated from measured values. BMIs based on self-reported height and weight values therefore underestimate the prevalence of overweight in adolescent populations.