This meta-analytic study examined the validity of Gordon's music aptitude tests as predictors of other musical variables. The four tests analyzed were the Music Aptitude Profile (1965), Primary Measures of Music Audiation (1979), Intermediate Measures of Music Audiation (1982), and Advanced Measures of Music Audiation (1989). Separate analyses were performed for tonal, rhythm, and composite constructs of music aptitude. From 47 music education journal articles that met requirements for inclusion, 215 independent data points representing 6,086 participants were collapsed into an overriding set of five criterion categories: (a) aural perception, (b) achievement, (c) creativity, (d) affective outcomes, and (e) musical engagement. Moderators potentially affecting validity included audiation type, sampling method, grade level, criterion test type, and year of publication. Results revealed estimated true criterion-related validities of 0.45 (tonal), 0.46 (rhythm), and 0.53 (composite). Gordon's music aptitude tests were consistently but not always strongly associated with many desirable musical outcomes. Analysis of correlations by subtest and criterion category produced mixed results, and high levels of between-study heterogeneity could not be explained through meta-regression moderator analysis.