Purpose. Reports from various pharmacy labor market sectors suggest that the United States may be experiencing a shortage of pharmacists. To guide policy making and planning with respect to this shortage it is necessary to develop a better understanding of the process by which pharmacists choose jobs. Using the economic theory of job matching, this study sought to understand how (a) attributes of the practice setting, (b) characteristics of pharmacists. and (c) regional and urbanization variables are associated with pharmacy practice setting choices. Methods. A secondary database containing information about employment characteristics and work histories of 541 pharmacists in four states was used. The data were augmented with information on the relative number of employment opportunities in each of three practice settings (large: chain, institutional, and independent) in the year the respondent's most recent employment change occurred. Practice setting choices were modeled using multinomial conditional logit regression. Results. A total of 477 pharmacists represented in the database met the inclusion criteria for the study. Multivariate analyses showed that the impact of search costs and wage differentials varied with the practice setting chosen. Pharmacists choosing independent settings over large chain settings were more likely to be white and to have worked in an independent setting in their prior job. Pharmacists living in Oregon were less likely to choose institutional settings compared to those living in Massachusetts, whereas those living in areas with populations greater than 50,000 were more likely to choose institutional settings. Conclusions. Pharmacist job matching appears to be a complex process in which diverse factors interact to produce a final match. Our results suggest that the: pharmacy labor market may actually be composed of two distinct labor markets: an ambulatory market and an institutional market.