The authors tested a 12-week parent training program with parents (n = 208) and teachers (n = 77) of 2-3-year-olds in day care centers serving low-income families of color in Chicago. Eleven centers were randomly assigned to 1 of 4 conditions: (a) parent and teacher training (PT + TT), (b) parent training (PT), (c) teacher training (TT), and (d) waiting list control (C). After controlling for parent stress, PT and PT + TT parents reported higher self-efficacy and less coercive discipline and were observed to have more positive behaviors than C and TT parents. Among toddlers in high-risk behavior problem groups, toddlers in the experimental conditions showed greater improvement than controls. Most effects were retained 1 year later. Benefits were greatest when parents directly received training.