Three studies had been completed and submitted for publication at the time of the interview. Researchers utilized a variety of data analysis methods: internal consistency of measures, means standard deviation of statistics, and discriminant function analysis with some follow-ups. The majority of participants (68.9%) used some form of assistive device (e.g., artificial limb, cane, crutches, scooter, wheelchair), were female (62.1%), and Caucasian (91.9%). The mean age of participants was 52.5 years (SD = 13.9). Participants completed and returned standardized questionnaires for each transtheoretical model (TTM) construct, along with a measure of exercise barriers. In univariate analyses, all of the constructs were significantly (p < .001) associated with the stages of change for exercise behavior. The largest portion of variance was derived from the behavioral processes of change (h2 = .40), followed by self-efficacy (h2 = .30), decisional balance (h2 = .24), the cognitive processes of change (h2 = .16), and exercise barriers (h2 = .15). Direct Discriminant Function (multivariate) analysis revealed three discriminant functions which accounted for 76.4% (p < .001), 19.4% p < .001), and 3.8% (p < .01), respectively, of the between-group (stage of change) variability. The overall stage of change classification accuracy was 67.4%. This was the first study to examine stage of change for exercise behavior among adults with physical disabilities on the basis of the full TTM and exercise barriers. Overall, the results are in general agreement with existing evidence among nondisabled populations. This provides further cross-sectional support for the internal validation of TTM, as well as the external validation of TTM and exercise barriers among a unique and understudied population segment. Behavioral strategies derived from TTM such as being moved emotionally, being rewarded, being a role model, developing a healthy self-image, gathering information, getting social support, making a commitment, making substitutions, taking advantage of social mores, and using cues may all facilitate adults with physical disabilities' stage of change for exercise behavior. While there could be some alternative interpretations of the data, the core findings are solid. Simple analyses are not as conservative. Discriminant functional analysis is more conservative. Using a univariate analysis provides more positive findings centering around the four or five most important functions. The data shows that behavioral processes of change are extremely important. Environmental factors (i.e. barriers) are also important.