Short and long-term retention in a visual recognition memory task was studied as a function of race using 39 10th grade inner-city high school students as subjects and random polygons as stimuli. It was hypothesized that a black subject confronted by an unfamiliar white adult and requested to take a test might be more aroused than a comparable white subject, and a related hypothesis was that all testing situations are more arousing to black than white subjects. A significant interaction of retention interval by race was obtained; for example, on the immediate test white subjects obtained higher recognition scores than black subjects while retention test recognition scores were higher for the black subjects on the one-week interval. These results supported an earlier hypothesis by F. H. Farley of arousal, memory, and the conditions of testing in black-white learning and memory research, suggesting in addition that the race of the experimenter may be crucial to the analysis of the black subjects' task performance. (Author/RB)