The purpose of this study was to examine the usefulness of individual load-velocity profiles and the between-athlete variation using the decrement in maximal velocity (Vdec) approach to prescribe training loads in resisted sled pulling in young athletes. Seventy high school, team sport, male athletes (age 16.7 ± 0.8 years) were recruited for the study. All participants performed one un-resisted and four resisted sled-pull sprints with incremental resistance of 20% BM. Maximal velocity was measured with a radar gun during each sprint and the load-velocity relationship established for each participant. A subset of 15 participants was used to examine the reliability of sled pulling on three separate occasions. For all individual participants, the load-velocity relationship was highly linear (r > 0.95). The slope of the load-velocity relationship was found to be reliable (coefficient of variation (CV) = 3.1%), with the loads that caused a decrement in velocity of 10, 25, 50, and 75% also found to be reliable (CVs = <5%). However, there was a large between-participant variation (95% confidence intervals (CIs)) in the load that caused a given Vdec, with loads of 14-21% body mass (% BM) causing a Vdec of 10%, 36-53% BM causing a Vdec of 25%, 71-107% BM causing a Vdec of 50%, and 107-160% BM causing a Vdec of 75%. The Vdec method can be reliably used to prescribe sled-pulling loads in young athletes, but practitioners should be aware that the load required to cause a given Vdec is highly individualized.