Over-harvest and landscape change are two of the greatest threats to marine ecosystems. Over-harvest may directly affect key population regulation mechanisms (e.g., density dependence), with the magnitude of the effects being further influenced by changes in landscape structure and associated resource availability. Because resource availability and conspecific density often co-vary within the natural landscape, manipulative experiments are needed to understand how changes in these two drivers may affect density dependence in wild populations. We used a common, shoaling, coral reef fish (white grunt, Haemulon plumierii) as our model species, and manipulated fish densities and landscape context of artificial reef habitats to assess the effects of each on fish condition. We found evidence of inverse density dependence, where individual condition was positively related to conspecific density; landscape context had little effect. Mean grunt condition on natural patch reefs was similar to that for our low grunt density treatment artificial reefs, possibly due to differences in fish densities or landscape context. These findings suggest that over-harvest may have detrimental effects on wild populations that extend beyond mere reductions in population size, especially for group-living species.