In response to the current shortage of rural physicians and the difficulties encountered in studying this problem, this paper attempts to apply a specific multivariate technique (path analysis) and the socioeconomic careers model of Featherman and others to the study of the physician's choice of practice location. The socioeconomic careers model is utilized as a heuristic device along with suggestions for its modification to make it more useful in the case of the physician. The specific classes of variables employed by Duncan and Blau are described in detail (background, intervening, career contingencies, and outcome variables). Certain of the specific aspects of the physicians' socioeconomic career are discussed. Differences between the generalized model and that found useful for the physician are pointed out (emphasis is on the use of many, rather than a few, variables). Consideration is given to discussion of path analysis as a type of multivariate analysis. The technique is then applied conceptually to the case of the physician's socioeconomic career. Specific variables are considered along with the conceptual difficulties they may present. Characteristics of actual recruits to medical careers are enumerated as indicative of how the model must be modified to be useful. (JC)