Background/Purpose: Identifying mediators of physical activity change requires measurement instruments that are reliable, valid, and generalizable to multiple populations. Despite continued application of the transtheoretical model (TTM) to the study of physical activity, the structural components of the TTM measurement instruments have been understudied in diverse populations. Methods: A multiethnic sample (N = 700, M age = 47, 63% women, 38% Caucasian) of participants living in Hawaii completed TTM measures. The factor validity and measurement equivalence/invariance (ME/I) of decisional balance, barrier self-efficacy, temptations, and processes of change instruments were explored between men, women, age groups, and ethnicities. Results/Conclusions: Measurement models of barrier self-efficacy and revised models of temptations and processes of change demonstrated sufficient evidence for ME/I among all subgroups. A revised model of decisional balance demonstrated sufficient evidence for ME/I between genders and among ethnicities, but not among age groups. Future research should examine the stability of these constructs across time. © 2008 The Society of Behavioral Medicine.