Background: Physical activity (PA) is important for management of diabetes, yet practical interventions that achieve sustained behavior change are rare.Purpose: The goals of this research were to evaluate the effectiveness of a multifaceted PA intervention for people with type 2 diabetes that emphasized participant choice in activity selection. Baseline activity patterns were examined to determine whether they predicted changes in PA at 2 months.Methods: Three hundred thirty-five participants were recruited from 42 primary care physicians and then randomized to either a computer-assisted, tailored self-management intervention (N = 174) or health risk appraisal with feedback control (N = 161). Primary outcome measures included the Community Healthy Activities Model Program for Seniors Questionnaire, diet, and psychosocial assessments at baseline and 2 months.Results: For 301 participants who completed the 2-month follow-up, the intervention significantly improved all PA (p < .01) and moderate PA (metabolic equivalents > 3.0, p < .01) relative to controls. Baseline cluster analyses grouped participant activity patterns into three categories. At 2 months, cluster assignment differentially predicted change in calories expended in moderate, rote, sport, and lifestyle PA.Conclusions: A computer-assisted, multifaceted approach to PA demonstrated improvement after 2 months. The results suggest that individuals are capable of adjusting their activity patterns to maximize their PA.