Zajonc's (1965) hypothesis was combined with a postulate of general drive theory to predict that the effects of anxiety and an audience would summate to increase the level of drive. Eighty Ss were used in a 2 by 2 by 2 design that had audience, anxiety, and list as factors. For one list competitive responses were high, and for the other they were low. Analysis of variance results and the order of the mean trials to criterion were consistent with the prediction based on the summation hypo thesis, e.g., of the eight conditions, the presence of an audience produced the best learning of the noncompetitive list and the poorest learning of the competitive list by high-anxious Ss. However, the lack of significant differences in the learning of the noncompetitive list did not support the summation hypothesis for this task.