Abstract Marine protected areas (MPAs) are increasingly integrated into fishery management for coastal systems. Size and spacing rules (SSRs) have been proposed as simple MPA design guidelines, especially in regions where population connectivity data are limited. We assessed whether SSRs allow managers to design effective MPA networks under spatiotemporally varying dispersal patterns using a spatially realistic population model parameterized for a commercially‐exploited fish species on the Great Barrier Reef. SSRs are used to design MPA networks, and population simulations are used to measure the mean and variance of the resulting population size and fishery catch. We show that SSR performance is contingent on the extent of the MPA network, and whether species’ connectivity data can be used to target areas for protection. For example, in the absence of connectivity data, a “many small” MPAs rule provides the least variable management outcome. Synthesis and applications. We demonstrate that the performance and usefulness of size and spacing rules (SSRs) as guidelines for marine protected areas (MPAs) depend on the level of knowledge about larval dispersal, as well as the level of current exploitation in the fishery. These context‐dependent results offer particularly relevant guidance to future MPA design projects in regions with limited connectivity data.