Thirty-six Ss arranged three sets of 15 words into easy serial orders. Approximately three weeks later they were required to learn three serial lists derived from the same sets of words. One serial list conformed to S’s own preferred order, one conformed to a preferred order developed by another S, and one was presented in a random order. Lists and presentation order of the three serial arrangements were counterbalanced. Serial learning was more rapid with easy orders, but there were only slight differences between the two preferred orders.