Patterns of bioerosion of dead corals and rubbles on the northern Great Barrier Reef were studied by using blocks of the massive coral Porites experimentally exposed at six sites, located on an inshore–offshore profile, for 1 year and 3 years. Rates of microbioerosion by microborers, grazing by fish, and macrobioerosion by filter-feeding organisms were simultaneously evaluated using image analysis. Microbioerosion, grazing, and total bioerosion were lower at reefs near the Queensland coast than at the edge of the continental shelf (1.81 kg m ?2 and 6.07 kg m ?2 after 3 years of exposure respectively, for total bioerosion). The opposite pattern was observed for macrobioerosion. Bioaccretion was negligible. These patterns were evident after 1 year of exposure, and became enhanced after 3 years. Microborers were established and were the main agent of bioerosion after 1 year of exposure, and as the principal support for grazing, continued to be the main cause of carbonate loss after 3 years. Full grazing activity and establishment of a mature community of macroborers required more than 1 year of exposure. After 1 year, macroborers and grazers were the second most important agents of bioerosion on both inshore and offshore reefs. However, after 3 years, grazers became the main agents at all sites except at the inshore sites, where macroborers were the principal agents. Because the contribution of microborers, grazers, and macroborers to bioerosion varies in space and time, we suggest that the estimation of reef carbonate budgets need to take in account the activities of all bioerosion agents.