This study tests the hypothesis that optimum testing style will result in more efficient test performance by 4-year-olds, thus diminishing differences between scores of lower class children who have and have not attended enrichment programs. A review of the literature discusses the effects of testing situation variables, language differences, dialects, ethnicity, motivational factors, and manipulation of pretest conditions on the scores of low income children. In the present study, scores on the Illinois Test of Psycholinguistic Ability (ITPA) were compared, since it was assumed that language scores were particularly likely to reflect the effects of optimum testing style. Data included the ITPA scores of (1) children participating in the Family Development Research Program (FDRP), an enrichment program for infants to school age children from low income families; (2) carefully matched control children who did not participate in the FDRP program but who were tested under FDRP conditions; and (3) children who had served as controls for other enrichment programs, in which testing was not done under FDRP conditions. Children in the FDRP experimental and control groups were not tested until they were at ease in the testing situation, with all attempts made to optimize test conditions. Although FDRP control children scored below children in the FDRP enrichment program, they performed only slightly below national norms on three of the ITPA subtests used in the comparison, and considerably above national norms on the other four subtests. Control children from the other enrichment programs generally performed below national norms on all subtests for which comparable data were available. Results and implications for testing and intervention programs are discussed. (SB)