Affective response to exercise may mediate the effects of self-paced exercise on exercise adherence. Fiftynine low-active (exercise <60 min/week), overweight (body mass index: 25.0–39.9) adults (ages 18–65) were randomly assigned to self-paced (but not to exceed 76% maximum heart rate) or prescribed moderate intensity exercise (64–76% maximum heart rate) in the context of otherwise identical 6-month print-based exercise promotion programs. Frequency and duration of exercise sessions and affective responses (good/bad) to exercise were assessed via ecological momentary assessment throughout the 6-month program. A regression-based mediation model was used to estimate (a) effects of experimental condition on affective response to exercise (path a = 0.20, SE = 0.28, f 2 = 0.02); (b) effects of affective response on duration/latency of the next exercise session (path b = 0.47, SE = 0.25, f 2 = 0.04); and (c) indirect effects of experimental condition on exercise outcomes via affective response (path ab = 0.11, SE = 0.06, f 2 = 0.10). Results provide modest preliminary support for a mediational pathway linking self-paced exercise, affective response, and exercise adherence.