This study investigated the importance of discipline variations in understanding faculty turnover behaviors. A representative sample of university faculty in Research and Doctoral universities was obtained from a national database. Faculty members, self-identified into a primary academic area, were grouped into eight discipline clusters according to an established framework. Multiple regression models were constructed to examine within each cluster the relative importance of a list of factors that have been identified to be related to faculty turnover. Cross-discipline comparisons of within-cluster variable prioritization revealed substantial discipline variations with regard to the major factors that are critical to faculty turnover. The findings produced evidence that discipline-specific information was indispensable to institutional administrators and policy makers for effective faculty retention.