Parks and protected areas have been instrumental in reducing anthropogenic sources of damage in terrestrial and aquatic environments. Pathogen invasion often succeeds physical wounding and injury, yet links between the reduction of damage and the moderation of disease have not been assessed. Here, we examine the utility of no-take marine reserves as tools for mitigating diseases that affect reef-building corals. We found that sites located within reserves had fourfold reductions in coral disease prevalence compared to non-reserve sites (80 466 corals surveyed). Of 31 explanatory variables assessed, coral damage and the abundance of derelict fishing line best explained differences in disease assemblages between reserves and non-reserves. Unexpectedly, we recorded significantly higher levels of disease, coral damage, and derelict fishing line in non-reserves with fishing gear restrictions than in those without gear restrictions. Fishers targeting stocks perceived to be less depleted, coupled with enhanced site access from immediately adjacent boat moorings, may explain these unexpected patterns. Significant correlations between the distance from mooring sites and prevalence values for a ciliate disease known to infest wounded tissue (r = −0.65), coral damage (r = −0.64), and the abundance of derelict fishing line (r = −0.85) corroborate this interpretation. This is the first study to link disease with recreational use intensity in a park, emphasizing the need to evaluate the placement of closures and their direct relationship to ecosystem health. Since corals are modular, ecological processes that govern reproductive and competitive fitness are frequently related to colony surface area therefore, even low levels of cumulative tissue loss from progressing diseases pose significant threats to reef coral persistence. Disease mitigation through reductions in physical injury in areas where human activities are concentrated is another mechanism by which protected areas may improve ecosystem resilience in a changing climate.