49 familial retarded children who had been tested on a simple 2-part satiation game 3 years earlier were retestcd on the same game. The children were divided into high and low preinstitutional social deprivation groups, and 2 reinforcement conditions were employed. On the original testing a positive relationship was found between social deprivation experienced and the effectiveness of social reinforcement (p<.05), with no significant reinforcement effects being found. On retesting, significant reinforcement condition effects were found (p<.OS), but no significant preinstitutional deprivation effects. An examination of the differences in the children's performance in the 2 testings revealed a significantly greater enhancement in the effectiveness of social reinforcement for the low than for the high socially deprived Ss (p<.01).