The study involved two stages: first, the evaluation of various stimuli as positive or negative; and second, the determination of whether or not positive stimuli were associated with the color white, and negative stimuli with black. The stimuli were statements related to self perception. From four integrated preschools were randomly selected 15 white males, 15 white females, 15 black males, and 15 black females. All were from middle to upper middle income families, with an average age of five years, nine months. The testing procedure involved first establishing the evaluation of each self statement by having the child point to either a painted smiling or frowning face; and second, broadcasting each statement to the child such that the sound originated with equal intensity from each of two recorder speakers, one painted white and the other black. The basic prediction (that children would "hear" the positive self statements coming from the white box and the negative self statements coming from the black box) was supported by the data, the data for white subjects more clearly supportive than that for black subjects, and more for males than females. (JM)