The study of self‐serving attributions in sports settings is fertile ground for testing the validity of self‐serving attributional phenomena. This paper reports the results of a meta‐analytic review of research examining self‐serving attributions in the context of sports events. A total of 91 distinct hypothesis tests were located, comprising five dimensions of attribution: ability (N= 21), effort (N=21), task difficulty (N= 21), luck (N= 21), and a general internal‐external dimension (N= 7). The meta‐analytic combination of significance levels indicated that the combined results were unlikely to occur if the null hypothesis of no effect were true (for each of the five dimensions of attribution). The internal‐external dimension and the ability dimension produced effects of moderate magnitude, whereas effort, difficulty and luck produced effects of small magnitude. Meta‐anaiytic focused comparisons revealed that self‐serving attributions (ended to be more extreme in the context of larger team sizes, and for attribution measures focused upon the team rather than the individual. Discussion considers the implications of these findings and develops and explanation for the finding that ability is the specific attribution dimension exhibiting the greatest self‐serving