<p>Scholars disagree on the distinction between positive and negative leader humor, which may hinder the integration and advancement of the leader humor literature. Thus, we meta-analyzed the correlation between positive and negative leader humor, as well as their shared nomological networks (<i>k</i> = 161; follower <i>N</i> = 45,759; leader <i>N</i> = 7820). Our meta-analysis indicates that (1) there is a very small but significant correlation between positive and negative leader humor (<i>ρ</i> = − .07); (2) positive and negative leader humor have differential relationships with most antecedents (e.g., positive affectivity) and outcomes (e.g., innovation and citizenship behavior); (3) positive affect and leader-member exchange intervene <i>independently</i> and <i>in series</i> the relationships between leader humor and follower outcomes, whereas negative affect mediates the effects of positive leader humor but not negative leader humor. These findings improve our understanding of the uniqueness of positive and negative leader humor, and provide important theoretical and practical implications for the leader humor literature.</p>