A series of experiments was carried out to examine the effect of moderate and maximal exercise on decision-making performance in team games. Subjects' speed and accuracy of decision making were examined using tachistoscopically presented tests of decision making in sports specific tests. The complexity of the decision making tests was manipulated as was the instructional set given to the subjects. Subjects were tested at rest and while exercising at 70% and 100% of their maximal power output (MPO). Multiple analysis techniques showed unequivocally that, regardless of complexity or instructional set, performance at rest was significantly poorer than that during maximal exercise. Observation of the data showed that this difference was mainly due to increases in speed of decision with accuracy making no significant contribution to the results. The results concerning the effect of moderate intensity exercise were equivocal. In some studies performance during exercise at 70% MPO was significantly better than that at rest, while in other studies no significant effect was demonstrated. Similar results were found when comparing performance during 70% and 100% MPO. In all experiments, however, a linear trend was observed. This was due to the contribution of changes in speed of decision, with accuracy making no significant contribution to the results. It was concluded that maximal exercise results in increases in speed of decision making in team games. This can probably be best explained by exercise inducing changes in arousal, which in tum leads to an increase in the available central nervous system levels of allocatable resources. It was further argued that maximal exercise does not stress fit subjects enough to cause the individual to allocate resources to anything other than task specific information. The use of theories of emotionally induced arousal to explain the effect of exercise on decision-making performance was questioned.