The present study was a meta-analysis of 29 published studies that used the Profile of Mood States (POMS) to investigate relationships between mood and athletic achievement or between mood and performance outcome. Results showed that effect sizes (ESs) for level of achievement were minimal (n = 15, Weighted Mean ES = 0.10, SD = 0.07), a finding consistent with a previous meta-analysis by Rowley, Landers, Kyllo, and Etnier (1995).Larger effects were found for performance outcome (n = 17, Weighted Mean ES = 0.31, SD = 0.12). Effects were moderate for Vigor, Confusion, and Depression, small for Anger and Tension, and very small for Fatigue. All effects were in the direction predicted by Morgan's (1985) mental health model. Effects were larger in sports of short duration, in sports involving open skills, and where performance was judged using self-referenced criteria.Findings suggest that the POMS has utility in the prediction of performance outcome but not in the prediction of level of achievement.Key words: Meta-analysis, Mood, POMS, Moderating variables The Profile of Mood States and Athletic Performance: A Meta-analysis There is a strong intuitive and anecdotal association between mood states and sport performance. However, empirical support for mood-performance relationships has been equivocal, despite the fact that more than 250 published studies have examined mood responses in sport and exercise settings (LeUnes & Burger, 1998) Sport psychology researchers have relied almost exclusively upon the Profile of Mood States (POMS: McNair, Lorr & Droppleman, 1971) as the measure of mood when examining links with athletic performance. The use of the POMS in sport was pioneered by Morgan and his co-workers (e.g. Morgan 1974;Morgan & Johnson, 1978;Morgan & Pollock, 1977;Nagle, Morgan, Hellickson, Serfass, & Alexander, 1975) who demonstrated that, when compared to population norms, the mood profiles of athletes particularly at the elite level were characterized by above average Vigor scores and below average scores for Tension, Depression, Anger, Fatigue, and Confusion. Morgan termed such a pattern of mood responses an iceberg profile and proposed that it was reflective of positive mental health (Morgan, 1980(Morgan, , 1985.Of the many specific research questions addressed within the area of mood and sport, three of the most frequently investigated have been (a) can mood responses differentiate the athlete from the non-athlete, (b) can mood responses differentiate athletes of varying levels of achievement, and (c) can mood responses differentiate performance outcome among athletes of similar ability? In respect of the first research question, qualitative reviews of the extant literature by LeUnes, Haywood, and Daiss (1988), Renger (1993), and Vanden Auweele, De Cuyper, Van Mele, and Rzewnicki (1993) demonstrate clearly that athletes typically report iceberg profiles, which by definition vary from population norms derived largely from nonathletes. Further, recently published normative data based on the mood res...