A length-structured simulation model is presented as a tool in decision making for gillnet fisheries management. The analytical model simulates the fish population dynamics and impact of the fisheries, taking into account size-dependent spatial distribution patterns and migration of the fish. The daily recruitment of fish, at a theoretical starting length of 0 cm, is simulated with a variable temporal pattern, giving the model a high degree of generality and making it applicable to tropical situations where recruitment is not restricted to a short period in time. The model is sex-differentiated, the growth of the fish is determined by their length and the dispersion over the length-classes is controlled by the 'fractional boxcar train method'. Natural mortality is assumed constant for all exploited size classes. Fishing mortality is a function of fish length, based on a combination of gillnet selectivity curves for various mesh sizes. The fishing mortality for the most vulnerable size class is calibrated using estimates for the fishing mortality per length-class from length-based cohort analysis. Spatial distribution of the fish over different habitats is based on observed patterns in size-dependent habitat occupation. The realistic model provides predictions on the progressive effects of management regulations on numbers, size distributions and total biomass of the population and the catch in the various habitats. The commercial gillnet fishery for Oreochromis mossambicus (Cichlidae) in Tissawewa, a Sri Lankan reservoir, is used as an example. Simulation results correspond with independently estimated values for output parameters. Based on model predictions, conservation of this fishery is recommended by re-enforcement of the legal minimum mesh size of 76 mm stretched mesh, and a limitation of fishing effort to the present level. The model also provides predictions on the effects of subsidiary small-meshed fisheries for minor cyprinids on the existing fishery for O. mossambicus. Model predictions take into account the size-dependent spatial distribution patterns of the fish and the spatial allocation of the fishing effort.