Genetic studies of population or 'stock' structure in exploited marine fishes typically are designed to determine whether geographic boundaries useful for conservation and management planning are identifiable. Implicit in many such studies is the notion that subpopulations or stocks, if they exist, have fixed territories with little or no gene exchange between them. Herein, we review our long-term genetic studies of red drum (Sciaenops ocellatus), an estuarine-dependent sciaenid fish in the Gulf of Mexico and western Atlantic Ocean. Significant differences in frequencies of mitochondrial DNA haplotypes and of alleles at nuclear-encoded microsatellites occur among red drum sampled across the northern Gulf of Mexico. The spatial distribution of the genetic variation, however, follows a pattern of isolation-by-distance consistent with the hypothesis that gene flow occurs among subpopulations and is an inverse (and continuous) function of geographic distance. However, successful reproduction and recruitment of red drum depend on estuarine habitats that have geographically discrete boundaries. We hypothesize that population structure in red drum follows a modified one-dimensional, linear stepping-stone model where gene exchange occurs primarily (but not exclusively) between adjacent bays and estuaries distributed linearly along the coastline. Gene flow does occur among estuaries that are not adjacent but probabilities of gene exchange decrease as a function of geographic distance. Implications of our hypothesis are discussed in terms of inferences drawn from patterns of isolation-by-distance and relative to conservation and management of estuarine-dependent species like red drum. Based on estimates of the ratio of genetic effective population size and census size in red drum, observed patterns of gene flow in red drum may play a significant role in recruitment.