Twenty-seven parents with young children were randomly assigned to an experimental group which underwent an affective skill-building program, or one of two control groups. Pre and postassessments measured levels of communication, discrimination, and child vocalization for each parent. Multilinear regression analysis indicated that final communication skills among the three groups were significantly different. Final communication skills of the experimental group were significantly greater than those of the Hawthorne control group. Final discrimination skills for the three groups showed a trend toward being significantly different. Levels of child vocalization did not show significant changes. The experimental program was successful in improving accurate parent-child communication in the affective realm.