The purpose of this investigation was to determine the relation between the affective value of color-combinations and that of the same colors experienced singly, with special reference to the relative size of the component colors. Procedure: Six saturated colors, red, orange, yellow, green, blue and purple were selected from Milton-Bradley papers for stimulus materials. The experiments were divided into two parts and the method of paired comparison was used throughout. In experiment I, six colors (9×9 cm. each) were presented two at a time in a row by means of an exposure-apparatus and the observer was asked to judge which was the more pleasant. In experiment II, a combination of two different colors, each in square and one within the other, wan used as a stimulus. According to color with the square outside (9×9 cm.), there were six series of stimuli consisting of 15 color-combinations each, as the square inside was of three different sizes (6×6 cm., 4.5×4.5tm. and 2.3×2.3 cm.) and all possible pairs of colors were combined. Comparison with respect to pleasantness was made only among color-combinations which belong to the same series and in addition have squares with the same color inside. Observer was seated 93 cm, distant from the exposure apparatus which was illuminated by two Mazda 100-watt day-light bulbs standing 75 cm. behind observer. Five observers (H. M., K. S., S. S., S. O. and K. M.) took part in these experiments. Results: 1. The affective rank order of 6 colors seemed to be a matter wholly of individual inclinaton. 2. It was not always the case that a combination of two pleasant colors was pleasant and that of two unpleasant ones unpleasant. It was also found that the traditional view regarding the pleasantness of complementaries in color combination was unfounded. 3. The affective value of a color combination did not re- main the same when the size of the square inside varied. No such simple relation was found, however, that the greater the area with the plesasant color in a combination, the greater was the pleasantness of the combination, 4. It may be said in conclusion that a simple summation of affective values of components, even if weighted by their areas, does not hold in color-combinations and the pleasantness of a combination is a function of combination itself, i. e., the balance between "powers" of the components.