The purpose of this review is to consider the prevalent types of fixed and flexible formularies, the general economic principles on which they are based and the evidence for their effectiveness in controlling rising drug expenditures. The principal-agent relationship and economic model underlying the various types of formularies are described. The principal-agent model describes a relationship where there is an asymmetry of information between two parties involved in a particular task. As a result of this asymmetry of information, the party with less information (the principal) allows the party with more information (the agent) to make decisions about that task or activity for them. In the case of formularies and cost-control, the principal is the payer. Depending on the incentives offered by the formulary, the agent can alternately be the prescriber, dispenser or patient. The success of a formulary type to control costs is dependent on two main factors. First, the payer (the principal) must identify the agent for whom it is reasonable to create incentives that incorporate the financial risks associated with use of the drugs. Second, the payer must develop a structure that best aligns the principal and agent objectives. The principal-agent framework serves as the vehicle through which the authors examine five major types of formularies (i.e., closed, best available price, reference-based pricing, tiered and open formularies) and their inherent incentives and limitations. The evidence for their effectiveness as cost-control mechanisms is reviewed and the system factors that can affect formulary success will be discussed. Finally, the authors' observations are summarized and interpreted, and suggested implications for future use of formularies in controlling the costs of pharmaceutical use are offered.