We present the first major test of proximal withdrawal states theory (PWST; Hom, Mitchell, Lee, & Griffeth, 2012). In addition, we develop and test new ideas to demonstrate how PWST improves our understanding and prediction of employee turnover. Across 2 studies, we corroborate that reluctant stayers (those who want to leave but have to stay) are similar to enthusiastic leavers (those who want to leave and can leave) in affective commitment, job satisfaction, and job embeddedness, and that reluctant leavers (those who want to stay but have to leave) are similar to enthusiastic stayers (those who want to stay and can stay) on these dimensions. We find that job satisfaction and job embeddedness more strongly influence the intent to leave and job search behavior for enthusiastic stayers and leavers than for reluctant stayers and leavers. More important, we show that for those experiencing low control over their preference for leaving or staying (i.e., reluctant stayers and leavers), traditional variables such as job satisfaction, job embeddedness, and intent to leave are poor predictors of their turnover behavior. We further demonstrate that focusing on enthusiastic stayers and leavers can significantly enhance the accuracy of job satisfaction, job embeddedness, and intent to leave for predicting actual employee turnover.