A comprehensive model of the work-family interface was developed and tested. The proposed model extended prior research by explicitly distinguishing between work interfering with family and family interfering with work. This distinction allowed testing of hypotheses concerning the unique antecedents and outcomes of both forms of work-family conflict and a reciprocal relationship between them. The influence of gender, race, and job type on the generalizability of the model was also examined. Data were obtained through household interviews with a random sample of 631 individuals. The model was tested with structural equation modeling techniques. Results were strongly supportive. In addition, although the model was invariant across gender and race, there were differences across blue- and white-collar workers. Implications for future research on the work-family interface are discussed.