This study examined the effectiveness of the Science Curriculum Improvement Study (SCIS) materials for visually handicapped children. Fourteen students, ranging in age from 9 to 19 and working in small groups with student teachers, received science instruction using the adapted version of the SCIS program. Control subjects were matched closely on age, medium of reading, socioeconomic status, and additional handicaps. These children participated as usual in the science program offered at their schools. Both groups were tested in the fall and the following spring. Three tests in the form of individual interviews in which each child was asked to examine and manipulate concrete objects, make hypotheses, and draw conclusions, provided the following results: Pretest results analyzed by means of the Mann Whitney U showed no significant differences between groups. Posttest data showed a superior performance by the experimental group over the control group on the three tests, "Histogram,""System and Interaction," and "Science Learning," significant at the .05, .025, and .001 levels respectively. The third test was broken into subtests of logic, manipulative, and science content items. Significant differences were found between the two groups on these subtests as well. (Author/EB)