Nine coral survey methods were compared at ten sites in various reef habitats with different levels of coral cover in Kāne'ohe Bay, O'ahu, Hawai'i. Mean estimated coverage at the different sites ranged from less than 10% cover to greater than 90% cover. The methods evaluated include line transects, various visual and photographic belt transects, video transects and visual estimates. At each site 25 m transect lines were laid out and secured. Observers skilled in each method measured coral cover at each site. The time required to run each transect, time required to process data and time to record the results were documented. Cost of hardware and software for each method was also tabulated. Results of this investigation indicate that all of the methods used provide a good first estimate of coral cover on a reef. However, there were differences between the methods in detecting the number of coral species. For example, the classic "quadrat" method allows close examination of small and cryptic coral species that are not detected by other methods such as the "towboard" surveys. The time, effort and cost involved with each method varied widely, and the suitability of each method for answering particular research questions in various environments was evaluated. Results of this study support the finding of three other comparison method studies conducted at various geographic locations throughout the world. Thus, coral cover measured by different methods can be legitimately combined or compared in many situations. The success of a recent modeling effort based on coral cover data consisting of observations taken in Hawai'i using the different methods supports this conclusion.