Compiled abundances of juvenile corals revealed no change over time in the Pacific, but a decline in the Caribbean. Using these analyses as a rationale, we explored recruitment and post-settlement success in determining coral cover using studies in the Caribbean (St John, Bonaire) and Pacific (Moorea, Okinawa). Juvenile corals, coral recruits, and coral cover have been censused in these locations for years, and the ratio of juvenile (J) to recruiting (R) corals was used to measure post-settlement success. In St John and Bonaire, coral cover was stable but different between studies, with the CSIRO PUBLISHING Marine and Freshwater Research, 2015, 66, 609-622 http://dx.doi.org/10.1071/MF14139 Journal compilation Ó CSIRO 2015 www.publish.csiro.au/journals/mfrratio of the density of juveniles to density of recruits (J : R) ,0.10; in Moorea, declines in coral cover were followed by recovery that was related to the density of juvenile corals 3 years before, with J : R ,0.40; and in Okinawa, a decline in coral cover in 1998 was followed by a slow recovery with J/R ,0.01. Coral cover was associated positively with juvenile corals in St John, and in Okinawa, the density of juvenile corals was associated positively with recruits the year before. J : R varied among studies, and standardised densities of juvenile corals declined in the Caribbean, but increased in the Pacific. These results suggest that differences in the post-settlement success may drive variation in coral community structure.