The purpose of this paper is to describe a specific method of measuring fathers' childrearing involvement. The conceptual scheme underlying the method addresses involvement in routine child care, play with the child, and school-related interactions. Measures involved the father's share of childrearing (as compared with the mother's) and the father's individual interaction frequency in each area of involvement. The sample, 70 non-black fathers of 6-year-olds, was drawn from the larger database of the Comparative Ecology of Human Development Project at Cornell University. Overall, among child development, socioeconomic, family demographic, and employment variables, the report card scores of male children were most related to measures of interaction frequency with the father. Father's reported interaction frequency with female children was not found to be positively related to report card scores. Family background variables appeared to be related primarily to the father-mother relationship. Fathers with greater education and white-collar occupations took a greater share of routine child care tasks and school-related responsibilities than did men with less education and blue-collar occupations. As compared with mothers, fathers with more children tended to take a smaller share in involvement with each child. Surprisingly, neither father's nor mother's paid work hours per week were related to the father's amount of involvement. The constructed measures were found to be reliable and valid. (RH)