The spatial dynamics of spawning fish are crucial because they influence the survival rates of eggs and larvae and ultimately impact the reproductive success of populations. The factors that control these dynamics are complex and potentially many, and they interact. A model-selection-based approach was developed to confront various hypotheses of control of the spatial distribution of spawning population of North Sea plaice (Pleuronectes platessa). For each hypothesis or combination thereof, statistical models were constructed. These were then ranked and selected based on their ability to adjust and predict observed spatial distributions. The North Sea plaice population seems to have developed strong attachment to specific spawning sites, where geographic location and population memory are important controlling factors. Temporal changes in spatial distribution patterns appear to be influenced primarily by population size and demography. Variations in hydrographic conditions such as temperature and salinity do not appear to control interannual fluctuations in spatial distribution. This means that, for reproduction, applying conventional habitat models may falsely attribute major controlling effects to environmental conditions. It is concluded that a multiple-hypothesis approach is essential to understanding and predicting the present and future distribution of the North Sea plaice population during its spawning season.