Coral reefs in many regions are under threat because of global warming, and local anthropogenic environmental changes and monitoring of the dynamics of coral communities is indispensable for understanding their response to these stresses. Different taxonomic resolutions have been used in the monitoring of coral communities depending on the research objective and the availability of resources. In this study, we examined the variation in the patterns of coral communities with the use of different taxonomic resolutions and assessed which of these were most sensitive to changes. Richness and composition measures of coral communities at 9 sites in Motobu, Okinawa, southwestern Japan, were monitored in 2006 and 2010. We compared taxonomic richness of corals among sites using univariate analysis, and taxonomic composition based from percent cover of each taxonomic group at different taxonomic resolutions using multivariate analysis. The pattern in terms of taxonomic composition employing life-form (taxonomic resolution based on colony morphology) were similar with the finer taxonomic resolution (family, genus, and “genus×life-form”) in 64% of all the possible site pair-wise comparisons in 2006, and 75% in 2010. Thus monitoring using life-form taxonomic resolution may be sufficient, particularly in the analyses of composition measures when resources are limited. However for the analyses of richness measurements, use of at least family level of taxonomic resolution may be recommended because the pattern of the richness difference among the sites at the life-form level were different from those at the finer levels of taxonomic resolution in 39% of the comparisons, especially in 2010. The adequacy of the use of coarser taxonomic resolution has some important implications for coral reef management because most monitoring is conducted by non experts in coral taxonomic identification.