The most common theme of previous turnover research is the attempt to predict turnover. However, the majority of previous turnover literature has ignored the dynamic or unfolding nature of turnover decisions. The present study re-evaluates the relationship between antecedents and turnover from a longitudinal approach. The longitudinal turnover approach incorporates change in the initial status and the slopes of the important turnover predictors over time as well as change in the nature or strength of the relationships between those variables and turnover risks over time. Two types of statistical analyses, survival analysis and growth modeling, are applied to assess questions that arise from the longitudinal turnover perspective, such as questions surrounding whether and when turnover occurs or questions surrounding the systematic changes of the relationship between predictors and turnover over time. The results of survival analyses indicate that psychological indicators, including employees' general attitudes towards the organization, their job satisfaction, and their intention to quit, have strong association with turnover risks over time. Management predictors, such as employees' compensation levels and their promotion history also have strong relations with turnover hazards over time. The results of growth modeling show that not only do initial levels of predictors have strong relationships with turnover risk, but so do their changing slopes. Overall, survival analyses and growth modeling analyses provide an opportunity for researchers to have a better understanding of the relations between predictors and turnover, longitudinally.