A large body of research on medical students' choice of primary care specialties has been published. However, the literature is difficult to interpret because of multiple biases, design weaknesses, small numbers of subjects, inconsistencies in both dependent and independent variables, and conflicting results. These weaknesses have been noted by authors who have reviewed the work in this area, but the authors have given little direction for ways to improve and build upon the current state of the literature. This paper provides a quantitative description of the content of the specialty choice literature. As part of a larger project that included an exhaustive literature analysis, all research on primary care specialty choice published between 1987 and 1993 was collected and summarized according to study questions, designs, data sources, samples, theory, and outcome variables. Portions of this information were used to rate the quality of each study, yielding a score from zero to 100 that indicated the trustworthiness of the study's conclusions. Overall, the studies examined were found to use predominantly cross-sectional designs and to lack theoretical basis. Special curricular tracks, student personality, and self-reported influences were the most frequently studied determinants of primary care specialty choices. The results confirm previous qualitative descriptions of the state of the literature on specialty choice, and lead to recommendations for approaches to improve the quality of further work in this area. The research agenda that emerged from the larger project is also presented.