According to motor-learning research, feedback is crucial to students’ development of motor skills. Based on this research, teacher feedback is considered an essential instructional strategy in physical education and is strongly emphasized in most teacher training programs (Graham, Holt/Hale, & Parker, 2001). The reality is most physical education students receive minimal amounts of individual skill performance feedback due to large classes. Feedback is more often used by teachers to provide general skill prescriptions to the class as a whole and as a class management tool. Due to these limitations, Rink (2003) suggests teachers must incorporate alternative methods for providing student feedback. The purpose of this study was to examine third grade students’ ability to accurately assess their performance on four dribbling tasks. Participants were 20 third grade physical education students at a rural elementary school. Students were videotaped performing four dribbling tasks (stationary dribbling, dribbling while moving, directional dribbling around cones, and guarded dribbling). Immediately following each task, students assessed their performance through either live or video recall on four skill cues (fingerpads, hand on top of the ball, head up and eyes looking forward, and ball below waist) using a checklist (always, sometimes, never, don’t know). One week later students performed the dribbling tasks again and assessed their performance using the alternate recall method. Two trained observers (interobserver reliability = 94 %) independently assessed the students’ performance, and scored the students’ responses as either correct, overestimated, or underestimated. Criterion results of the dribbling tasks showed students struggled with accurately assessing the dribbling cues. The “hand on top of the ball” cue (M = 71.3 % correct) was the easiest for students to judge. The cues of “fingerpads” (M = 66.2), “ball below waist” (M = 62.7), and “head up and eyes looking forward” (M = 51.9) showed decreasing levels of success. Students were more accurately able to assess performance from video recall (M = 65.9) than live recall (M = 61.2). Boys (M = 68.1) more accurately assessed skill performance than girls (M = 58.6), although both were more likely to overestimate than underestimate. Overall, the results indicate the students in this study were minimally successfully at assessing their own skill performance during one episode of dribbling regardless of the type of recall. These results suggest that specific training may be needed to help students accurately assess skill performance.Keyword(s): assessment, curriculum development, elementary education