Many coral reefs throughout the world continue to decline in health and resilience. In the Caribbean, much of this decline has been attributed to increases in algal cover due to the loss of herbivores such as herbivorous fishes, especially large parrotfish, and the sea urchin Diadema antillarum. In recent years, the combination of protections and natural recovery has helped these herbivores, particularly parrotfishes, to recover in some areas. This recovery has led to questions about how different herbivores affect both the overall benthic cover of macroalgae and the specific composition of algal communities. Toward that end, we designed a caging experiment in Little Cayman to manipulate density and type of grazers. Treatments included full cages (1.27 cm wire mesh) with 4 D. antillarum/m2 added, full cages with 1 D. antillarum/m2, lidless cages (fences) open to herbivorous fishes without D. antillarum, full cages (herbivore exclosures) that excluded most fishes and urchins and open cageless controls. Benthic cover was quantified to assess differences in percent cover and community structure of macroalgae and benthic invertebrates throughout the 5-month study. As the experiment proceeded, both percent algal cover and algal community structure in the full herbivore exclosures, treatments open to fishes (open plots and fences), and treatments with D. antillarum (both densities) diverged. No changes were observed in percent cover of corals or other benthic invertebrates. Overall, reductions in percent algal cover were greater in the high density (4/m2) D. antillarum treatment compared to the low density (1/m2) D. antillarum treatment or those treatments where fishes were allowed access. As expected, greater densities of sea urchins reduced algal cover more quickly. However, these high density treatments also affected the D. antillarum themselves. The condition of the urchins (assessed as ratio of mass to test diameter) at the conclusion of the experiment declined in the 4 D. antillarum/m2 treatment compared to urchins in the 1 D. antillarum/m2 treatment and natural populations. Moreover, high densities of D. antillarum were better able to reduce the abundance of the brown macroalgae Lobophora spp. and Dictyota spp. compared to natural assemblages of herbivorous fishes. These findings suggest that these different types of herbivores, i.e., fishes and D. antillarum, exert different pressures on available algae. Their relative abundances will likely dictate macroalgal community structure on Caribbean coral reefs, and thus the suitability of a reef for the settlement of vital sessile organisms, especially corals.