Several quantitative models of coalition formation assume that a coalition is successful if the strength of the coalition is greater than the strength of the target, but unsuccessful otherwise. However, strong empirical evidence in favour of this hypothesis is still lacking. In this study, we provide an empirical test of this assumption in Barbary macaque males, by using a field-based estimate of individual competitive ability from which coalition strength is derived. Coalition success was determined for 90 coalitions composed of two partners and targeted at one male. Of these, 72.2% were behaviourally successful and 27.8% were unsuccessful. Asymmetry in strength was a significant predictor of coalition success, as this factor alone could explain up to 78.6% of coalition outcomes in the study group. Males behaved as if they were at least partially informed about the nature of this asymmetry. The targets of attacks by coalitions were more likely to counterattack as asymmetry in strength decreased, and coalition partners formed coalitions that produced on average a greater asymmetry in strength than would be expected by chance. However, we provide evidence that males may have used simple rules of thumb based on their knowledge of dyadic and third-party relationships, rather than estimates of asymmetry in strength per se. We conclude that competitive ability is an important factor in coalition formation in Barbary macaque males and discuss additional factors not included in this study, which may account for the unexplained outcomes.