Background: The purpose of this study was to evaluate the accuracy of timing lights (TL) at different heights for measuring velocities during sprinting. Methods: Two sets of single beam TL were used to determine velocities reached in a flying 20-m sprint in 15 healthy and physically active male participants. In TL64, all TL were set up at a height of 64 cm, and in TL100, all TL were set up at 100 cm, respectively. Participants performed three valid trials. The recordings of high-speed video cameras were used as a reference. Results: ICC and Pearson’s r values between both timing light heights and the reference system were almost perfect (0.969–0.991). Bland & Altman’s LOA (95 %) indicated low systematic and unsystematic errors, with somewhat smaller LOA for TL100 (-0.013–0.121 m/s) than for TL64 (-0.060–0.120 m/s). Measures of between-trial reliability of running velocities showed a high relative (ICC) and absolute (RMSE) reliability, with the reference system showing slightly better values in all reliability measures (ICC=0.935; RMSE<0.001 m/s) compared to TL64 and TL100 (ICC=0.894, 0.887; RMSE=0.107 m/s, 0.124 m/s, respecti ... mehrvely). The usefulness, determined by comparing the typical error (TE) with the smallest worthwhile change (SWC), was considered as “OK” (TE ≈ SWC) for all three systems. Conclusions: Results suggest that TL at both heights (TL64 and TL100) can be considered as accurate, reliable, and useful in computing velocities during a flying 20-m sprint, and therefore can be recommended to both coaches and researchers.