Seagrass fishes perform key ecological roles and are a critical component of many of the world's fisheries. A sound understanding of seagrass fish communities, based on robust methods, is therefore integral for their successful environmental management and conservation. Here we quantitatively compare taxa and size class of seagrass fish assemblages collected using; baited and unbaited remote underwater video (BRUV, URUV), diver operated video (DOV), underwater visual census (UVC) and trawling. We assess comparisons of these methods in terms of biodiversity indicators and power to detect change. All five methods were deployed in coastal embayments dominated by Posidonia spp. and within three marine parks along the south-western coast of Western Australia. Fish assemblages recorded using the different methods separated into two distinct groups; those dominated by smaller and less motile species, that are typically present beneath the canopy among seagrass leaves; and those dominated by larger, faster swimming species present above the seagrass canopy. Trawling provided the most effective method for assessing species within the canopy, including those of conservation concern. Conversely, BRUV followed by URUV recorded the greatest number of individuals, successfully detecting the larger supra-canopy and highly motile species typically omitted by trawling. Both trawl and BRUV methods had a higher power to detect change in species richness of seagrass fishes compared to other methods, even though each method recorded different components of the fish assemblage. Fish assemblages recorded using UVC detected both inter and supra-canopy species, but fish abundances recorded by this method were low compared to trawl and BRUV surveys, resulting in reduced power to detect change in richness. DOV was a comparatively less effective method, recording few fish, as well as low richness and diversity values. The combination of trawling and stationary video techniques, particularly BRUV, provided a two-method combination able to holistically measure commonly used indicators of fish assemblage composition in seagrass habitat. These results demonstrate that careful consideration of the methodology is essential when assessing ichthyofauna in seagrass habitat, particularly when only a single method can be employed. As the extent of methodological influence on recording fish assemblages also varied among the three survey areas, geographical differences in the structure of fish and seagrass communities should be considered when selecting an appropriate sampling method.