In 1967, seven demonstration kindergarten classes were set up in order to observe children's characteristics; to determine activities, individualized instruction, and teacher aide functions and training; to develop curriculum and plans for evaluation and management routines; to examine instructional materials and class size; and to help parents understand the functions of kindergarten. The population included 320 children, seven teachers, seven teacher aides, and seven principals. Record data were collected, and instruments which could be cross-checked were used. Reliability was achieved by pretesting and posttesting with standardized tests. Teachers and aides were oriented toprocedures before evaluation, and highly competent personnel were employed. Conclusions reached were as follows: (1) kindergarten programs should be flexible to meet varying cognitive, social, emotional, and physical needs of the children; (2) evaluative instrument" should be developed and giv"n priority to assess programs; (3) demonstration centers should be organized and inservice training should be established for prospective teachers; (4) teaching strategies and curriculum should be studied in depth; and (5) home-school relations and role of all personnel should be delineated. Followup studies on all phases of the program were recommended. (DO)