Though the processes of change are conceived as the core constructs of the transtheoretical model (TTM), few researchers have examined their construct validity in the physical activity domain. Further, only 1 study was designed to investigate the processes of change in an adolescent sample. The purpose of this study was to examine the exercise behavior processes of change in a sample of adolescents using novel three-step confirmatory validation procedures, including (a) the item-level aggregation of processes, (b) the higher order structure of aggregation, and (c) the discriminant validity of processes across the stages of change. Participants were 15 to 17 year old high school students (N = 284) who completed measures of the exercise behavior processes of change and stages of behavior change. Results using structural equation modeling identified that the processes of Dramatic Relief, Helping Relationships, Environmental Reevaluation, and Self-Reevaluation had acceptable item-level measurement properties (p less than 0.01), while the construct of Social Liberation was not supported. Distinct processes structures were a better fit of the observed data than any higher order behavioral or experiential structure (p less than 0.01). Finally, validation tests of the processes of change across stage of change using discriminant function analysis and univariate F tests supported Counter Conditioning as the critical process of exercise behavior change in adolescents (p less than 0.01) and found no support for the experiential processes. Changes and implications to TTM theory and measurement were discussed as well as practical implications for exercise behavior intervention strategies. (Contains 6 tables and 1 figure.)