The performance subscales to the Wechsler Intelligence Scale for Children were administered to 208 male and 208 female junior high and high school students between 12 and 16 years of age. The variables of test atmosphere (evaluative or gamelike), tester expectation (high or low), race of tester (black or white), and race of subject (black or white) were placed in a 2 x 2 x 2 x 2 factorial design. The pattern of mean IQ scores as well as self-ratings of mood were interpreted as indicating that test performance was optimal at moderate levels of motivational arousal. A replication of the experiment employing 208 male subjects increased cell sizes to the point that socioeconomic status could be treated as an independent variable. The results suggested that interracial differences in mean IQ might be erased depending upon the social psychological characteristics of the test setting and the socioeconomic background of the testee.When an IQ test is administered to children attending a racially integrated school, blacks generally average from 10 to 20 points below whites, though there is customarily a substantial overlap in the two distributions (Jensen, 1969). As Sattler (1970) has pointed out, one approach toward identifying the origins of such interracial differences in mean IQ is to examine the social psychology of the test situation itself. Since black students have in the past usually been tested by white examiners under conditions of ego threat, for instance, some investigators have proposed that fear of failure on the one hand and fear of appearing "uppity" on the other may each impair the performance of blacks in this milieu (Katz & Cohen, 1962;Sarason, Davidson, Lighthall, Waite, & Ruebush, 1960). Other researchers have found direct evidence that the achieve-