The purpose of this study was to examine the influence of different tasks on behavior in mother-child interactions. Five tasks, varying in degree of structure and difficulty, were used in mothers' interactions with their children at 24 and 36 months of age. Subjects were Mexican-American mothers and their children. All were participants in an extensive 2-year parent education program designed especially to meet the needs of low-income Mexican-American families. Families entered the project when the child was one year of age. Participants were randomly assigned to program or control groups and then each mother-child interaction was videotaped and rated at one minute intervals on the following scales: Affection, Praise, Criticism, Control, Reasoning, Mother's Verbal Encouragement, and Child Verbal Responsiveness. The Bayley Mental Development Index (MDI) was used to measure the children's abilities at 24 months of age and Palmer's Concept Familiarity Index (CFI) and the Stanford-Binet (S-B) were used when the children were 36 months of age. Results showed that: (1) program and control mother-child dyads differed significantly on mean scores for Affection, Criticism, Child Verbalization, and Mother's Verbal Encouragement and that with the exception of Criticism, all differences favored the program group at both time intervals; (2) while significant correlations were found between mother-child interaction variables and measures of child cognitive competence, there were no distinct differences between tasks in these relationships; and (3) for both groups, the correlates with measures of child competence were similar for high and low structure tasks while the relative stability of high and low structure tasks across time was also similar. (Author/MP)