Aim: The great variation in range sizes among species has fascinated ecologists for decades. In reef-associated fish species, which live in fragmented habitats and adopt a wide range of dispersal strategies, we may expect species with greater dispersal ability to spread over larger ranges. However, empirical evidence for such a positive relationship between dispersal and range size in reef fishes remains scarce. Here, we unveil the more nuanced role of dispersal on the range size distribution of reef associated fishes using empirical data and a novel spatially explicit model. Location: Tropical Eastern Pacific Major taxa studied: Reef-associated fishes Methods: We estimated range size distributions for six different guilds of all reef-associated fishes with different dispersal abilities. We used a one-dimensional spatially explicit neutral model, which simulates the distribution of species along a linear coastline to explored the effect of dispersal, speciation and sampling on the distribution of range sizes. Our model adopts a more realistic gradual speciation process (protracted speciation) and incorporates important long distance dispersal events with a fat-tail dispersal kernel. We simulated our model using a highly efficient coalescence approach, which guarantees the metacommunity, is sampled at dynamic equilibrium. We fitted the model to the empirical data using an approximate Bayesian computation approach, with a sequential Monte Carlo algorithm. Results: Stochastic birth, death, speciation and dispersal events alone can accurately explain empirical range size distributions for six different guilds of tropical, reef-associated fishes. Variation in range size distributions among guilds are explained purely by differences in dispersal ability with the best dispersers covering larger ranges. Main conclusions: A simple combination of neutral processes with guild-specific dispersal ability provides a general explanation for both within- and across-guild range size variation. Our results support the theoretically expected, but empirically much debated, hypothesis that dispersal promotes range size.