Wheelchair sports and daily manual wheelchair propulsion are dominated by frequent short-term power demands. The purpose of the current cross-sectional study was to determine the variation in propulsion technique in association with sprint power production among elite wheelchair athletes. Therefore, 67 wheelchair athletes (different impairments; 17 female and 50 male athletes; age, 29.1+/-7 yr; body weight, 60.7+/-11.8 kg; training hours, 12.9+/-6.4 h x wk(-1); VO2 peak, 1.7+/-0.7 liter x min(-1); aerobic power output, 72.2+/-36.7 W) were studied during the World Championships and Games for the Disabled in Assen (1990) on propulsion technique and anaerobic work capacity in a 30-s sprint test on a computer controlled wheelchair ergometer. Mean power output (P30) (97+/-45.8 W; range, 8.3-195.3 W) and heart rate (158.6+/-23.6 b x min(-1)) were highly variable and seemed associated with impairment level: track athletes, classified in four different functional classes, showed a class-related P30 of 23, 68, 100, and 138 W for the male athletes (n=38). Sprint power relative to body weight varied between 0.36 W X kg BW(-1)+/-0.04 and 1.85 W X kg BW(-1)+/-0.43 for the different subject groups. Propulsion technique in terms of forces applied to the rim and timing showed significant differences between subject groups for the majority of parameters studied. Apart from the mediolateral force and the negative dip at the start of the push phase, the technique parameters were significantly related to power production. Fraction effective force, the ratio between the total force vector and the effective force applied to the hand rim, appeared low on average (especially for subjects with cerebral palsy and those with a high spinal lesion) but showed a significant correlation with power output (r=0.5). In general, propulsion technique parameters were related to both performance and functionality. The number of training hours showed a small but significant relation with peak power (r=0.31), peak torque (r=0.4), the amount of work per push (r=0.41) and the total force vector (r=0.31), stressing the role of training status, next to disability, as important mediating factor in both propulsion technique as well as performance capacity. No association between training hours and fraction effective force was seen. It can be concluded that propulsion technique and performance parameters are highly variable among wheelchair athletes. Also, propulsion technique is strongly associated with functionality and training hours and does clearly relate to performance. The current results on technique and performance and their possible causal relationship, but also with impairment and sports discipline, must be further substantiated in a longitudinal study design.