Bycatch of marine megafauna in fishing gear is a problem with global implications. Bycatch rates can be difficult to quantify, especially in countries where there are limited data on the abundance and distribution of coastal marine mammals, the distribution and intensity of fishing effort, and coincident interactions, and limited bycatch mitigation strategies. The dugong Dugong dugon is an IUCN-listed Vulnerable species found from the eastern coast of Africa to the western Pacific. As foragers of seagrass, they are highly susceptible to bycatch in small-scale fisheries. To address the knowledge gaps surrounding marine mammal bycatch, we used existing survey and fishing effort data to spatially characterize the risk of bycatch for this species. Using Sabah, Malaysia, as a case study, we employed presence-only modeling techniques to identify habitat associations of dugongs using a maximum entropy distribution model (MaxEnt) based on published sightings data and several geophysical parameters: coastal distance, depth, insolation, and topographic openness. Model outputs showed distance from the coast as the highest-contributing variable to the probability of dugong presence. Results were combined with previously published fishing effort maps of this area to develop a predictive bycatch risk surface. Our analyses identified several areas of high risk where dugong surveys were conducted, but also identified high-risk areas in previously unsurveyed locations. Such methods can be used to direct field activities and data collection efforts and provide a robust template for how existing sightings and fishing effort data can be used to facilitate conservation action in data-limited regions.