Fish predation is an important disturbance affecting corals, yet few studies have compared the response of multiple coral species to one or more modes of fish predation. The present study differs from previous work on this topic by utilizing short-term field experiments, and testing the effects of corallivory on life stages (i.e., small or juvenile colonies) that are thought to be sensitive with regards to the effects of damage, and which play important roles in population growth. Our objective was to assess the growth response of two morphologically distinct coral taxa (massive Porites sp. and Pocillopora meandrina) to artificial damage designed to mimic the effects of three corallivorous fish feeding groups: excavators, scrapers, and browsers. The experiment was conducted in the lagoon of Moorea, and corals in each damage group (as well as controls) were allocated to replicate cages for 21 d to measure growth in response to treatment. The damage type significantly affected the growth of Porites and P. meandrina, with excavating reducing growth by 39% and 54%, scraping by 24% and 43%, and browsing by 13% and 17%, respectively. Excavation created lesions varying in size from 80-114 mm2 at the start of the study, and this created the opportunity to explore the effects of lesion size on growth. While it had no effect on P. meandrina growth, the growth of Porites was positively associated with lesion area, with more severe damage resulting in a stronger growth response. This counterintuitive result for Porites might reflect the effects of tissue biomass intertwined with a porous skeleton, but regardless of mechanism, the outcome appears to reduce the impacts of severe damage. Together, these results suggest that corallivory affects the success of small corals over a short time scale, with the effect depending on the type of damage. Thus, the relative abundance of fishes in each feeding group is likely to determine the overall impact of corallivory on coral populations.