In this thesis the nature of the market for physicians' services has been examined and the implications of its characteristics have been analyzed for the three basic systems of payment: fee-for-service, capitation, and salary. Because of the fact that the services provided are not homogeneous, the presence of external economies, and the informational inequality between patient and doctor, the market diverges significantly from that of the perfectly competitive model. A model was developed for the determination of the physician's equilibrium income under a fee-for-service system. Further analysis showed that the doctor will act as a market divider and will practice price discrimination among individual patients in the specialist market. His ability to affect the level of demand for his services leads to the greatest increase in his welfare when he performs inefficiently with respect to society as a whole, performing only those services whose price is greater than his estimate of the satisfaction he would derive in the future if the service were not performed. As a result, the price will increase and this price increase will spread through the system as a whole. The physician will innovate if and only if the cost or innovation is less than or equal to the expected increase in his income minus the amount of income necessary to compensate for the expected increase in disutility involved. If a third party provides part or all of the remuneration, the effects of the doctor's ability to induce demand are greatly magnified. A second model was developed to show the physician's equilibrium income under a capitation system. In this system, the quantity of services demanded is only minimally related to the price, and the doctor's marginal revenue for performing any given service is zero with respect to the present income period and patient list. Some limits must be set on the patients' ability to change doctors. Pressure also arises for a classification of patients and graduated levels of capitation payments; the system tends to become more closely related to a fee-for-service system. In addition, some patients must be assigned to physicians, and enforcement of minimum standards of care may be necessary if competition among doctors is weak. Referral of patients and use of related health facilities will be encouraged unless specific limits are placed upon them. Finally, the salaried system was analyzed in relationship to the capitation and fee-for-service systems. The salaried system alone necessarily involves an employer-employee relationship, and development of bargaining strength becomes the dominant financial goal of the doctor. His ability to induce demand will be used as it was in the capitation system. The incentive to utilize related services, to perform research, to innovate, and, indeed, to perform with respect to almost all other criteria will depend upon the emphasis placed upon these items by the employer. Because of differences in the perspectives of the physician, the third party, and the patient, a salaried system implies the need for machinery to resolve differences among the three perspectives. The three systems were then compared with respect to the risks and the accounting costs involved. For the physician, the level of total income becomes more certain with prepayment of fees, still more so with capitation payments, and most of all with a salary. In contrast, as this becomes more certain, uncertainty as to the amount of disutility undergone for any particular level of income is increased. Minimal accounting costs are incurred in a salaried system, and these costs increase for capitation, and especially for fee-for-service. Total cost may be made predictable within each system, but only with greater administrative costs in the capitation and especially the fee-for-service systems.