The purpose of the study was to evaluate two physical activity recall instruments appropriate for use in epidemiologic studies of fifth grade children. The instruments were similar, except one (PACI) was administered in a personal interview, and the other (SAPAC) was self-completed in a group setting. Both forms required children to report the minutes during the previous day they spent in 21 common physical activities that represented a range of intensities, plus sedentary pursuits. To validate the recalls, children simultaneously wore an accelerometer (motion sensor) and a heart rate monitor for at least 8 h the day before the interview. Subjects were 55 boys and 70 girls from four regions of the United States. The Pearson correlation between the self- and interviewer-administered forms was 0.76 (P < 0.001). The interviewer-administered form correlated 0.51 (P < 0.001) with the heart rate index and 0.33 (P < 0.001) with the accelerometer score. The self-administered form correlated 0.57 (P < 0.001) with the heart rate index and 0.30 (P < 0.001) with the accelerometer score. It is concluded that both self-report forms received moderate support for their validity in all gender and ethnic subgroups. The self-administered format is more cost-effective.