Recently, similar six-factor solutions have emerged in lexical studies across languages, giving rise to the HEXACO model of personality. As a core extension of its most well-known predecessor, the five-factor model, the HEXACO model distinguishes between two factors predicting complimentary aspects of prosocial behavior or, more specifically, reciprocal altruism: Honesty-Humility (the tendency toward active cooperation, i.e. non-exploitation) and Agreeableness (the tendency toward reactive cooperation, i.e. non-retaliation). However, this dissociation has not yet been tested to its full extent. To this end, we herein present re-analyses of published studies (N= 1090), showing that Honesty-Humility, but not Agreeableness, indeed predicts active cooperation. More importantly, in a new experiment (N= 410), we found a pattern of two concurrent selective associations, supporting the theoretical distinction between the two factors: Honesty-Humility (but not Agreeableness) predicted active cooperation (non-exploitation in the dictator game), whereas Agreeableness (but not Honesty-Humility) was linked to reactive cooperation (non-retaliation in the ultimatum game).