Although the effect of external load on the mechanical output of individual muscle has been well documented, the literature still provides conflicting evidence regarding whether the optimum loading (L(opt)) for exerting the maximum muscle power output (MPO) could be different for individuals with different levels of strength and power. The aim of this study was to explore the effect of training history on L(opt) that maximizes MPO during the 6-s maximal cycling sprint test. Forty healthy young males (strength-and speed-trained athletes, and physically active and sedentary non-athletes) were tested on maximum strength, and on peak MPO when loaded 5-12% of body weight (BW). As expected, the strength trained and sedentary participants, respectively, revealed the highest and lowest strengths and MPO (p < 0.001). However, the main finding was a significant across-group difference in L(opt) (p < 0.001) revealing the values 9.7% (for strength trained), 9.2% (speed trained), 8.7% (active), and 8.0% of BW (sedentary individuals). This suggests that the effects of external loading on maximum MPO in complex functional movements could be training history dependent. In addition to revealing a sensitivity of the 6-s maximal cycling sprint tests (and, perhaps, other maximum cycling tests), the results suggest that the external loading in routine MPO tests should not be solely adjusted to a fixed percentage of subject's BW (as routinely done in standard tests), but also to their training history. The same phenomenon remains to be evaluated in a number of other routine tests of MPO and other maximum performance tasks.