OBJECTIVE: The authors explored whether demographic and psychosocial variables predicted differences in physical activity for participants with arthritis in a trial of Active Living Every Day (ALED). METHOD: Participants (N = 280) from 17 community sites were randomized into ALED or usual care. The authors assessed participant demographic characteristics, self-efficacy, outcome expectations, pain, fatigue, and depressive symptoms at baseline and physical activity frequency at 20-wk follow-up. They conducted linear regression with interaction terms (Baseline Characteristic × Randomization Group). RESULTS: Being female (p ≤ .05), less depressed (p ≤ .05), or younger (p ≤ .10) was associated with more frequent posttest physical activity for ALED participants than for those with usual care. Higher education was associated with more physical activity for both ALED and usual-care groups. DISCUSSION: ALED was particularly effective for female, younger, and less depressed participants. Further research should determine whether modifications could produce better outcomes in other subgroups.