In this study, we established a motor skills test for college badminton courses. Subsequent physical education testing and educators in related fields can reference these results. We used repeated measures to collect data on 139 students in a badminton course at a national university. Two weeks before the end of the semester, we randomly selected students to test backhand drop shot service in badminton with 5 trials, 10 trials, 15 trials, and 20 trials. We analyzed these data with descriptive statistics, t-tests and product moment correlation. The results indicated that the backhand shot service skill test had content validity, test-retest reliability, and discrimination capability. After performing an oriented analysis on the standard deviations of the total scores, scoring rates, and average score per trial of the four types of tests with different numbers of trials, we found that 10 trials was the most appropriate number of trials, and sex did not influence the testing results. Therefore, we suggest 10 trials as a benchmark to conserve human labor and time and to reduce the influence of the fatigue caused by excessive numbers of trials.