The efficacy of three group process variables-group potency, social cohesion, and team-member exchange-in explaining team performance was examined for 648 military officers working in 50 self-managed teams over a 5-week period. Team performance was assessed on two objective criteria (mental task performance and physical task performance) and one subjective criterion (commander team performance ratings). Each of the group process variables was related to at least one aspect of team performance. Of the three group process variables, group potency exhibited the greatest predictive efficacy as it explained unique variance in team performance over that accounted for by both social cohesion and team-member exchange.