Marine protected areas (MPA) and in particular no-take marine reserves have been identified as important tools for the conservation of reef fish and habitats. A significant challenge of reef fish monitoring is to determine the influence of temporal factors on fish counts. Fish assemblages are dynamic and changes in activity patterns throughout the day can influence the results of surveys. While many monitoring programs account for the impacts of spatial heterogeneity on fish assemblages the effects of short-term temporal variation are less well known. In the present study, we analysed data from 197 video drops inside and outside New Caledonian MPAs and examined temporal variations in various metrics commonly used to monitor marine reserves. In addition to describing short-term temporal patterns related to time of day, tide height and state, and lunar cycle; we also examined the influence of these temporal factors on the size and direction of any MPA effects detected. Fewer Serranidae were observed at full moon than the rest of the lunar cycle and there were negative correlations between Chaetodontidae and Acanthuridae abundance and tide height. We did not find any consistent effects of time of day. Generally, variation in short term temporal factors did not affect the direction MPA effects detected but did affect the size of the effects for some metrics. For both small fish abundance and species richness, bigger differences between protected and unprotected sites were detected at high tide than low tide. These results highlight how survey results can vary with timing of sampling and have implications for developing optimal monitoring programs.