The knowledge that regular running provides many positive benefits has prompted many schools to establish running programs; however, little is known about children's motivation in such running programs. This study examined children's motivations in an elementary physical education running program using achievement goal theory and an expectancy-value model of achievement choice as theoretical frameworks. Specifically, the relationships among children's achievement goals (task-involved, ego-involved, and work avoidance), expectancy-related beliefs (ERBs), subjective task values (STVs) and their running performance and intention for future participation in running along with possible gender differences were examined. Participants were 119 fourth graders (67 boys; 52 girls) who participated in a running program conducted during their regularly scheduled physical education classes. A 28-item, 5-point scale questionnaire adapted from previous work with elementary children assessed the children's achievement goals, expectancy-related beliefs, subjective task values, and intention for future participation in running. A timed mile run assessed children's running performance. Correlation analyses revealed that the task-involved goal (p < .001) was positively related to ERBs and STVs. The ego-involved goal (p < .01) was positively related to the work avoidance goal and ERBs. The ERBs (p < .001) were positively related to STVs. All motivational variables (ps < .001), except the work avoidance goal, were positively related to children's intention for future participation in running and their running performance. The work avoidance goal was negatively related to children's running performance (p < .001) and ERBs (p < .01). Results of stepwise multiple regression analyses indicated that STVs (p < .001) were the only positive significant predictor for children’s intention for future participation in running, explaining 50% of the variance. For children's running performance, ERBs (p < .001) and ego-involved goal (p < .01) were significant positive predictors, while the work avoidance goal (p < .01) was a negative predictor. They accounted for 28.66% of the variance. A MANOVA revealed no gender effect, indicating that boys and girls did not score differently on the mean scores of the seven variables. Overall, our findings reveal that achievement goals, expectancy-related beliefs, and subjective task values are related to one another and are predictive of children’s intention for future participation in running and their performance on the mile run. We therefore advocate the use of multiple theoretical approaches for a comprehensive understanding of student motivation and achievement in physical education.