The purpose of this study was to summarize findings from empirical applications of the transtheoretical model (TTM) (Prochaska & DiClemente, 1983) in the physical activity domain by using the quantitative method of meta-analysis. Ninety-one independent samples from 71 published reports were located that present empirical data on at least one core construct of the TTM applied to exercise and physical activity. In general, results support the application because core constructs differ across stages and most changes are in the direction predicted by the theory. Three general conclusions are offered. First, existing data are unable to confirm whether physical activity behavior change occurs in a series ofstages that are qualitatively different or along adjacent segments of an underlying continuum. Second, the growing number of studies that incorporate TTM concepts means that there is an increasing need to standardize and improve the reliability of measurement. Finally, the role ofprocesses of change needs reexamining because the higher order constructs are not apparent in the physical activity domain and stage-by-process interactions are not evident. There now are sufficient data to confirm that stage membership is associated with different levels of physical activity, self-efficacy, pros and cons, and processes of change. Further studies that simply stage participants or examine cross-sectional differences between core constructs of the TTM are of limited use. Future research should examine the moderators and mediators of stage transition.