The study examined the longitudinal association of parental involvement in Head Start parent-focused programs, parenting behaviors, and the cognitive development of children by specifying two longitudinal growth models. Model 1 examined the longitudinal effects of the parental involvement in three Head Start parenting programs (parenting classes, group socialization, and support groups) on parenting behaviors (home observation of parental linguistic and cognitive stimulation, video recordings of parental cognitive stimulation, parental supportiveness, detachment, and intrusiveness). Model 2 analyzed the longitudinal effects of those parental behaviors on children's Bayley MDI scores. Using Early Head Start Research and Evaluation (EHSRE) study data and longitudinal multilevel analysis, the study also took various ethnic and language differences among families into account. The results revealed that mothers who participated in parenting classes or socialization meetings provided more linguistic and cognitive stimulation at home. Participants of parental support groups were found to have high levels of parental supportiveness and low levels of parental intrusiveness over time. Higher Barley MDI scores were found for children whose mothers had high levels of parental involvement in Head Start parent programs and provided more at-home linguistic and cognitive stimulation. The African American families, in particular, benefited from attending socialization meetings: attendees displayed fewer parental detachment behaviors and provided more linguistic and cognitive stimulation, resulting in higher Bayley MDI scores of children. The study's findings are theoretically significant and policy relevant. (Contains 3 tables.)