Purpose - The current study seeks to argue that the constructs of work demand and family demand have been neglected in the work-family conflict (WFC) literature. The authors aim to help clarify the definition and utilize direct measures of perceived work and family demand to test main effect, mediated, and interactive hypotheses. Design/methodology/approach - A sample of 698 university employees participated in a comprehensive computer survey that considered various manifest indicators and multiple scales across work and family domains. Moderator hierarchical regression and LISREL 8.0 were used in analyzing the data. Findings - The results indicate that both forms of demand have significant direct effects on work interfering with family (WIF) and family interfering with work (FIW). Both demand constructs partially mediate the effects of three categories of domain variables on the two forms of conflict. Finally, the work demand-WIF relationship is found to be stronger for those with relatively high family centrality. Research limitations/implications - A cross-sectional design was used and may be problematic when examining relationships that occur over time. Further, capturing all scales with a single survey could result in common method bias, which may have inflated the predictive relationships. Practical implications - Organizations can work to reduce WFC by adopting family-friendly programs that help employees balance work and family demands. Specifically, this study implies that organizations should find ways to hold constant or reduce perceptions of work and family demand, along with other direct antecedents of WIF and FIW. Originality/value - This study provides a relatively comprehensive model of antecedents that can be useful in future research. The authors also examine interactive effects of demand and work-family centrality on conflict using direct measures of perceived demand. Methodologically, the research improves on some past studies by measuring perceived demand directly and by not narrowing our sample to employees who are married or those with children. Hopefully, these contributions will help stimulate continued growth in the work-family literature.