The purpose of this study was to determine the test-retest reliability of a newly developed anaerobic sprint running test (ASRT) on a nonmotorized treadmill (NMT). Twenty-six collegiate male athletes (21.2 ± 2.1 years; 181.3 ± 6.5 cm; 79.0 ± 9.3 kg) completed 3 trials of a 25-second maximal effort sprint on an NMT against a workload set to 18% of their individual body mass. Anaerobic power was determined by relative peak power output (PP) and anaerobic capacity was determined by relative mean power output (MP) during the test. Blood lactate (BLa) responses and fatigue index (FI) were also determined. Test-retest reliability was assessed by intraclass correlation coefficients (ICCs) and coefficients of variation (CV%). Results indicated no significant difference between the 3 trials for PP (T1 = 29.95 ± 6.51 W·kg(-1), T2 = 28.57 ± 5.55 W·kg(-1), T3 = 29.47 ± 5.94 W·kg(-1)), MP (T1 = 20.97 ± 3.64 W·kg(-1), T2 = 20.50 ± 3.46 W·kg(-1), T3 = 21.17 ± 3.79 W·kg(-1), and FI (T1 = 55 ± 8%, T2 = 51 ± 8%, T3 = 52 ± 9%). Reliability between the 3 trials for PP (ICC: r = 0.96, CV: 7%) and MP (ICC: r = 0.97, CV: 6%) was considered high. Reliability for FI exhibited an ICC of r = 0.83 (CV: 6%). Postsprint BLa values were not significantly different (p = 0.49) between the 3 trials. Test-retest reliability for postsprint BLa was found to be good (r = 0.68, CV = 8.8%). The results of the study indicate that the ASRT is reliable for assessing PP and MP in highly motivated subjects. In addition, anaerobic testing using the ASRT may be a more sport-specific test to assess anaerobic performance for many coaches and athletes.