The theory underlying the design of marine reserves, whether the goal is to preserve biodiversity or to manage fisheries, is still in its infancy. For both of these goals, there is a need for general principles on which to base marine reserve design, and because of the paucity of empirical experience, these principles must be based on models. However, most of the theoretical studies to date have been specific to a single situation, with few attempts to deduce general principles. Here we attempt to distill existing results into general principles useful to designers of marine reserves. To answer the question of how fishery management using reserves compares to conventional management, we provide two principles: (1) the effect of reserves on yield per recruit is similar to increasing the age of first capture, and (2) the effect of reserves on yield is similar to reducing effort. Another two principles answer the question of how to design reserve configurations so that species with movement in various stages will be sustainable: (3) higher juvenile and adult movement lowers sustainability of reserves for biodiversity, but an intermediate level of adult movement is required for reserves for fishery management, and (4) longer larval dispersal distance requires larger reserves for sustainability. These principles provide general guidelines for design, and attention to them will allow more rapid progress in future modeling studies. Whether populations or communities will persist under any specific reserve design is uncertain, and we suggest ways of dealing with that uncertainty.