The current anthropogenic impacts on nature necessitate more research for nature conservation and restoration purposes. To answer ecological and conservation questions concerning endangered species, individual-based modelling is an obvious choice. Individual-based models can provide reliable results that may be used to predict the effects of different future conservation strategies, once calibrated correctly. Here, we calibrate an individual-based model of Maui dolphin movement, which generates Maui dolphin probability distribution maps. We used sighting data for calibration of the chosen parameter combinations; for each simulation run, collected simulated data was compared to the empirical survey data, resulting in cost (Badness-of-Fit) estimates. Using costs of four different aspects of dolphin behaviour, we estimated the most likely parameter combinations. With optimized parameter values, Maui dolphin probability distribution maps were created, resulting in distributions that fall well outside of the current protection zones where either gillnets or trawling or both are prohibited. With these results, protected areas can be properly adjusted to the estimated distribution of this critically endangered species and so aid in their conservation.