Although proposed as an explanation for increases in positive and decreases in negative affect, little evidence supports the notion of a stronger efficacy-affect relationship as acute aerobic exercise intensity increases. Relationships between self-efficacy (SE), positive affect (PA), and negative affect (NA) were examined with respect to 3 randomly assigned conditions: (a) no exercise (control), (b) cycling at 55% VO2max, and (c) cycling at 70% VO2max. Twenty subjects (age = 22.6 years; M VO2max= 47.8 ml · kg−1· min−1) participated in each. Preexercise SE predicted in-task NA and postexercise PA in the 55% condition and postexercise PA in the 70% condition (ps < .05). Although SE significantly increased from pre- to postexercise, in-task affect failed to predict these increases in SE. These data suggest that in a fit college-aged population, a stronger reciprocal relationship between preexercise SE, in-task affect, and postexercise SE does not exist with increasing levels of exercise intensity. Contrary to self-efficacy theory, no evidence was found for a reciprocal relationship. However, the relationship between efficacy and affect was such that when collapsed across conditions and time, subjects with higher SE scores reported lower NA (p < .05).