The object of this study was to investigate the critical attack and defense skills that influenced the victory or defeat in intercollegiate men's handball competitions, and to establish a regression formula to evaluate the victory or defeat in intercollegiate men's handball competitions. In accordance to the Intercollegiate Handball Competitions of 1992 thru 1997, the difference of points gained and lost recorded from the men's A and B groups, were used as criterion for the Mallows Cp and R-square regression analysis. Moreover, the cross validation was calculated from records of the men's A and B group results at the "2000 South-East Asia Intercollegiate Handball Invitational Tournament" held at Shan-Tou University, and the "2002 Intercollegiate Handball Championship". Fouls (X1), slow break fumbling (X2), fast break fumbling (X3), slow break block by the goal-keeper (X4), penalty throw blocked by the goal-keeper (X6), block by a court player (X7), field shooting (X8) and fast break shooting (X12) were chosen among 19 variables, obtaining an R-square of 0.6280 (F=55.93, p<.05) and the following regression formula to evaluate victory or defeat in intercollegiate men’s handball competitions: ŷ=5.76039-0.3416(X1)-0.69894(X2)-0.5666(X3)+0.7442(X4)+1.5003(X6)+0.5350(X7)-0.4910(X8)+0.5345(X12) The validation obtained was R^2=0.5643, with an estimated standard error of 1.8203, and an external validity merely reduced by 0.0637. Consequently, the conclusion of this study was that the critical attack and defense skills that were chosen were the optimum combination. Furthermore, they possessed a good external validity and evaluation quality.