The purpose of this study was to examine the relationship between brain activity and mood state for varying combinations of exercise intensity and duration.  Male subjects (N = 32) were recruited to participate in a no exercise condition and in six randomly presented combinations of two intensities (50-55% & 70-75% of VO2 max) and three durations (15, 30, & 45 min.) of treadmill running.  Mood was assessed by the Activation-Deactivation Adjective Checklist and the short version of the STAI.  EEG and these instruments were administered before and at 5- and 15 min post-exercise.  At 15 min postexercise in the 15 min of high intensity exercise condition, the results showed that the proportion of general activation (GA) variance accounted for by the pre-exercise resting frontal EEG asymmetric index over and above the pre-exercise GA was 5% (P < 0.05).  It is concluded that the resting frontal asymmetric index predicted post-exercise affect in only one of the 51 instances where pre- to post-exercise mood changes were observed.  Because this finding does not represent more than would be expected by chance alone, the present data do not support Davidson's idea that the asymmetric index predicts subsequent affect changes only when a sufficient emotion-eliciting stimulus is presented.  In this study, exercise was an emotion-eliciting stimulus for many of the combinations of exercise intensity and duration.  However, with only one exception, the EEG asymmetric index did not predict post-exercise affect where significant pre- to post-exercise mood changes occurred.