The purpose of the present study was to define the optimal loads (OL) for eliciting maximal power-outputs (PO) in the leg and arm modes of the 30s Wingate Anaerobic Test (WAnT). Eighteen female and seventeen male physical education students, respectively 20.6 +/- 1.6 and 24.1 +/- 2.5 years old, volunteered to participate. In each of the total five sessions, the test was administered twice on a convertible, mechanically braked cycle-ergometer, once for the legs and once for the arms. The five randomized, evenly-spaced resistance loads ranged from 2.43 to 5.39 Joule per pedal revolution per kg body weight (B. W.) for the legs, and from 1.96 to 3.92 for the arms. The measured variables were mean (MP x kg-1) and peak PO as well as absolute and relative measures of fatigue. A parabola-fitting technique was employed to define the optimal loads from the MP x kg-1 data. The resulting OL were 5.04 and 5.13 Joule x Rev-1 x kg B.W.-1 in the leg and 2.82 and 3.52 in the arm tests for the women and men, respectively. OL were shown to depend on PO magnitude. However, within a two-load span (0.98 Joule x Rev-1 x kg B.W.-1) about the OL, MP x kg-1 did not vary by more than 1.4% in the leg and 2.2% in the arm tests. It is suggested that although the WAnT is rather insensitive to moderate variation in load assignment, improved results could be obtained by using the stated OL as guidelines that may be modified according to individual body build, composition, and, particularly, anaerobic fitness level.