Reservoirs are subject to artificial water level fluctuations leading to a degradation of the littoral habitats. This often results in reduced productivity and biological diversity, particularly for fish. The development of predictive tools of fish population dynamics linked to habitat characteristics can help to optimize management actions such as habitat restoration. We developed a spatially explicit individual-based model, the 3Plake model, which predicts the spatial distribution of three fish species, northern pike, European perch and pikeperch, over seasons and diel cycles according to their habitat use. The model was calibrated with an accurate description of habitat and high spatial and temporal resolution data of fish movements in a French reservoir. The model calibration is only based on a species-specific coefficient of habitat selectivity. It was chosen to (i) match the observed and simulated traveled hourly and daily distances and (ii) minimize a metric quantifying the distance between observed and simulated density maps. The 3Plake model correctly simulated the means of traveled distances and the species distributions along the year on the study reservoir. The minimal input required (HSI maps and maximum distance range) allows its easy implementation on other reservoirs and fish species. It could be used to assess the impact of habitat restoration or water level management scenarios on fish distribution.