AS PART OF AN EFFORT TO ASSESS THE EFFECT OF THE BALTIMORE HEAD START PROGRAM ON COGNITIVE DEVELOPMENT, CHILDREN ENROLLED IN THIS PROGRAM WERE GIVEN THE PEABODY PICTURE VOCABULARY TEST (PPVT) AND THE DRAW-A-PERSON (DAP) TEST AT THE START OF THE PROGRAM, AT ITS TERMINATION, AND WHEN THEY ENROLLED IN KINDERGARTEN. A CONTROL GROUP, DRAWN FROM THE SAME NEIGHBORHOODS, WAS TESTED AT KINDERGARTEN ENROLLMENT. THE NUMBER OF CHILDREN TESTED IN EACH GROUP WAS SLIGHTLY OVER 400. DATA ON THE FAMILIES OF THE HEAD START CHILDREN INDICATED THAT THEY WERE SEVERELY DISADVANTAGED. RESULTS OF THE PPVT SHOWED SIGNIFICANT GAINS DURING THE PROGRAM AND BETWEEN THE END OF THE PROGRAM AND KINDERGARTEN ENROLLMENT. THE CONTROL GROUP SCORED AT THE SAME LEVEL THAT THE HEAD START CHILDREN HAD SCORED AT THE BEGINNING OF THE PROGRAM. RESULTS OF THE DAP TEST SHOWED A SIMILAR PATTERN OF GAINS, BUT WITH THE CONTROL GROUP SCORING AT THE SAME LEVEL THAT THE HEAD START CHILDREN HAD SCORED AT THE END OF THE PROGRAM. CLASSROOM EVALUATIONS OF TEACHER BEHAVIOR WERE MADE, AND PRELIMINARY ANALYSES INDICATE THAT TEACHERS WHO WERE RATED AS WARM, VARIED, AND FLEXIBLE AND WHO SPENT A RELATIVELY HIGH PROPORTION OF TIME TEACHING PRODUCED THE HIGHEST GAINS. (DR)