80 Ss learned a serial list of 10 CVCs under four experimental conditions. Two rates of presentation (fast and slow) and two isolation conditions (isolation and no-isolation of the sixth item) were combined in a 2 × 2 factorial design. A significant interaction of isolation and rate of presentation was found. This supports the hypothesis derived from “differentiation” theory (Saltz, 1960) that the relative effect of isolation would be greater in a list with a little degree of differentiation (fast rate of presentation) than in a more differentiated one. Two additional hypotheses based on this theory were tested and discussed.