From a sample of 681 college undergraduates 4 groups of 20Ss of each sex (M, F ) , comparable in age and reasoning ability, were selected for low (L = 3 to 9 ) and high (H = 27 to 41) anxiety on the basis of Taylor MAS scores. All SO Ss received 320 trials on the Air Force Discrimination Reaction Timer wirh nonspecific instructions designed to produce a high degree of task complexity. The results were as follows: ( 1 ) Group HF was slightly inferior in reaction speed ( R . ) to Group LF initially, superior later; ( 2 ) Group HM remained markedly inferior to Group LM throughout; ( 3 ) men were generally faster than women; (4) women were significantly more anxious pre-experimentally than men; ( 5 ) the main effects of Practice, Sex, and their interaction were significant, as predicted, but Anxiety was a clear decremental factor only among males; ( 6 ) mean acquisition curves for 4 augmented groups of 50 Ss each were exponential and also revealed a marginal Practice X Sex X Anxiety interaction.The equations fitted the R, data wirh an average error of less than 196, which is in excellent agreement wirh the Hull-Spence theory. Implications for prior research by Grice, Noble, Spence, and Taylor were discussed.A considerable amount of research has been done on human performance as a function of variations in motivation employing extensions of Hull's (1943) general behavior theory. Useful summaries of the literature may be found in the writings of Brown (1961), Farber (1954), Spence (1956, 1958, and Tay lor ( 1956). Investigations in this area have dealt with the role of nonspecific drive factors in learning situations where level of motivation was varied both experimentally and psychometrically.As applied to latency measures, the core postulates of the Hull-Spence theory assert chat for single-response learning casks: 'This paper is based upon a master's thesis performed by J. for checking the starisrical analyses.