In two experiments, performance on a modified version of the Stroop color-word task varied systematically with level of task-irrelevant arousal. Performance under low response competition was facilitated while performance under high response competition was impaired by arousal manipulated via threat of impending electric shock. The present results were thus consistent with traditional theory relating arousal or drive and response competition and suggest that Stroop task performance may thereby provide a behavioral index of arousal level.While performance on the Stroop color-word task (Stroop, 1935) has been widely used as a index of various individual differences (cf. Jensen & Rohwer, 1966), recently Hochman (1967, 1969 and Pallak and Pittman (1972) suggested that performance on a modified version of the task may also provide an index of arousal level. For example, within traditional learning theory, motivational constructs such as drive facilitate responses both instrumental and noninstrumental to reduction of drive level (cf. Hull, 1943;. In task situations in which habit strength of the correct response is dominant or has a high probability of emission, an increment in drive level facilitates the response and thereby enhances task performance. However, in task situations in which habit strength of the correct response is low, that is, where habit strength of an incorrect response is high, an increment in drive level facilitates the competing response and thereby diminishes successful task performance (cf. ).Hochman 's (1967, 1969) version of the Stroop task varied verbal response competition by serially presenting a list of nouns printed in various colors of ink and requiring the subject to name the color of the ink. In the high response competition conditions, the nouns employed named competing colors. In the low response competition conditions, the nouns employed had no strong color associations. Stress or arousal was manipulated by varying the length of response time permitted fpr the subject to name the color of the ink for each noun-ink color stimulus item. Subjects in the high arousal condition were allowed I sec, while subjects in the low arousal condition were allowed 2 sec to This research was supported in part by a faculty research grant to the first author awarded by the Graduate College (University of Iowa) from National Institutes of Health Biomedical Sciences Research Program funds and in part by traineeships to the second and third authors from National Institutes of Health Training Grant funds in the azea of social psYchology. The second, third, and fourth authors are now at GettYsburg College, Franklin and Marshall College, and University of Kansas, respectively. This paper was sponsored by Rudolph W. Schulz who takes full editorial responsibility. respond to each presentation. Hochman (1967Hochman ( , 1969 found that high stress resulted in more verbal errors in the high response competition version but failed to fmd that high stress resulted in fewer errors under low response competition r...