Abstract A grid-based, spatially-explicit Regional Organism Exchange (ROE) model is presented as a framework for integrating aquatic ecosystems and fish population processes at the landscape level. Active fish movements across a grid cell boundary were predicted, based on environmental tolerance ranges. The model was designed to be easily modified for any aquatic system, migratory life-stage, or trophic community. ROE was specifically developed to understand how large-scale physical patterns (i.e., tidal and freshwater intrusions) and landscape biological processes (i.e., primary production and foraging behavior) control migration of stenohaline fishes in the estuarine lagoon of Laguna de Terminos, Mexico. A migration response matrix for temperature, salinity, food availability, birth, and mortality was used to control cell-to-cell population movements. Internal cell processes included logistic population growth, trophic interactions, and ecosystem feedback parameters. Output data maps from the ROE model showed how population spatial distributions were linked to spatial and temporal patterns of water quality. However, the most significant parameter affecting long-term population stability was birth rate; an internal cell variable. It was concluded that the simulation of large, density-dependent, spatial processes such as migration can be understood with a grid-based mechanistic ROE model because its rule-based design for movement allowed organisms to respond to ecological processes and adjust to changing environmental conditions.