The second year of a state supported early childhood education program was evaluated, using a pre-post evaluation design involving experimental and control groups. An Early Childhood Assessment Battery was administered to 720 five-year-olds enrolled in the kindergarten program by their teachers. The control group (178) were tested locally. Results include: (1) The steadiness of the trends in non-cognitive areas from the beginning to the end of the year, and from one year to the next, gives confidence that the Classroom Behavior Inventory is a means for measuring non-cognitive aspects of children's school experiences; (2) Helping increase children's behavior in the areas of the positive behavior traits--and decrease the negative--has a positive effect upon cognitive performances; (3) In cognitive changes, the experimental group exceeded the control group in absolute scores on every test; (4) All children who attended kindergarten the year before are maintaining a reasonable average with respect to national norms; and (5) The children who had attended kindergarten the year before continued to make substantial gains between the beginning and ending of first grade on all measures. (Author/CK)