This dissertation compares three of the leading theories on wage and promotion dynamics, tournament theory, job assignment theory, and signaling theory, via their predictions on the impact of promotion on employees' turnover behavior. In the first chapter entitled "Theory on Promotion and Turnover", I incorporate turnover into all three theories through an employeremployee specific match factor and generate three predictions from each theory. The first two are similar among the theories and the third distinguishes among them. Specifically, all the theories predict that promotions are positively correlated with wage levels and that performance ratings are positively related to the probability of promotion. Regarding promotion and turnover dynamics, the tournament model predicts that promotion is negatively related to turnover, while the job assignment model predicts that neither promotion nor the wage is related to turnover. Further, under certain conditions, the signaling model predicts that promotion is positively related to turnover and that the wage is negatively related to turnover controlling for promotions. In the second chapter entitled "Theoretical Extensions on Promotion and Turnover," I further extend both tournament and job assignment theories by adding firm specific human capital. These extensions enrich the predictions of both theories regarding promotion and turnover dynamics. Specifically, the further extended tournament theory predicts that promotion is negatively correlated with turnover while the wage is negatively correlated with turnover controlling for promotions. On the other hand, the further extended job assignment theory predicts that promotion is positively correlated with turnover while the wage is negatively correlated with turnover given promotions. In the third chapter of my dissertation entitled "Evidence on Promotion and Turnover," I test these predictions using a dataset on employees from a single firm in the financial services industry in the U.S. I first show that tournament theory better captures the promotion-turnover dynamics in the full sample. I then conduct further tests by breaking up the sample into a high level subsample and a low level subsample. I find that tournament theory better explains the promotion-turnover dynamics in higher level jobs, while both signaling theory and job assignment theory with firm specific human capital better explain the dynamics in lower level jobs. I interpret this finding as evidence consistent with both an incentive argument related to Rosen (1982) and Baker, Jensen, and Murphy (1988) and the hybrid approach discussed in Waldman (2013).