Weighted sled towing is used by athletes to improve sprint acceleration ability. The typical coaching recommendation is to use relatively light loads, as excessively heavy loads are hypothesized to disrupt running mechanics and be detrimental to sprint performance. However, this coaching recommendation has not been empirically tested. This study compared the effects of weighted sled towing with 2 different external loads on sprint acceleration ability. Twenty-one physically active men were randomly allocated to heavy- (n = 10) or light-load weighted sled towing (n = 11) groups. All subjects participated in 2 training sessions per week for 8 weeks. The subjects in the heavy and light groups performed weighted sled towing using external loads that reduced sprint velocity by approximately 30 and 10%, respectively. Before and after the training, the subjects performed a 10-m sprint test, in which split time was measured at 5 and 10 m from the start. The heavy group significantly improved both the 5- and 10-m sprint time by 5.7 ± 5.7 and 5.0 ± 3.5%, respectively (P < 0.05), whereas only 10-m sprint time was improved significantly by 3.0 ± 3.5% (P < 0.05) in the light group. No significant differences were found between the groups in the changes in 5-m and 10-m sprint time from pre- to posttraining. These results question the notion that training loads that induce greater than 10% reduction in sprint velocity would negatively affect sprint performance and point out the potential benefit of using a heavier load for weighted sled towing.