Several series of stimuli, varying in shape, in color, or in both, were presented according to the method of pair comparisons to separate groups of infants 4 mo of age. Relative visual fixation time served as a measure of preference. Shape was quantified in terms of contour, number of turns, or symmetry. Color was designated according to Munsell coordinates. It was found that: (a) red and blue were significantly preferred to gray, (b) a bulls-eye pattern was preferred to other patterns, and (c) shape dominated color as the basis for preferential choice when stimuli represented combinations of preferred and non-preferred colors and shapes as determined for other experimental gr oups.