Cardinal scores of individual competitive ability allow us to quantify the magnitude of the difference between the competitive ability of any two individuals. However, they have rarely been used in animal behaviour because most researchers were mainly interested in ordinal ranking. In this paper, we validated the normalized David’s score (David H. A., Biometrika 74, 1987, 432; de Vries H., Stevens J. M. G. & Vervaecke H., Anim. Behav. 71, 2006, 585) as a cardinal measure of male competitive ability in a group of Barbary macaques living under semi-free ranging conditions. To derive competitive ability scores, we used a semi-experimental protocol where two males had to compete over access to a prized food resource (i.e. a nut) within the natural group setting. This protocol was used because it allowed the exclusion of three factors other than competitive ability (i.e. respect of ownership, social tolerance and motivation) which may influence the outcome of dyadic encounters in group-living primates. We expected that a measure of competitive ability excluding the three above-mentioned influences would correlate with some intrinsic features of males. Male competitive ability scores were calculated based on 357 nut tests. As expected, male competitive ability shows a curvilinear relationship with age (used as a proxy for male general physical condition), with young, ‘athletic’ males having the highest scores. However, we also found that male competitive ability scores were highly correlated with the dominance scores derived from naturally occurring agonistic interactions, which suggest that observations of spontaneous interactions may suffice to estimate the competitive abilities of individuals. We conclude that despite its limitation, the normalized David score is often preferred to ordinal ranking as an estimate of resource holding potential as originally defined (Parker G. A., J. Theor. Biol. 47, 1974, 223).