Information for monitoring marine communities is increasingly collected using digital imagery and there is a need to understand how data collected using different methods compare. This study investigated the comparability of benthic data collected by downward facing diver operated stereo-video (DOV) with a forward facing stereo-DOV typically used in fish surveys. Data sets were collected by both methods from the same transect along which the benthos visible in each method was identified by taxonomy, growth form or substratum class. Analyses indicated that 77% of taxa were identified by both methods however, forward facing-DOV recorded 35 genera and downward-DOV 43 genera and there were significant (p < 0.05) differences in the percent cover of dominant benthic categories. The forward facing stereo-DOV detected significantly more vertically erect coral (∼ 10%), specifically branching genera Acropora, Echinopora, and Porites, and canopy forming algae, Sargassum spp., but proportionally less turf algae (∼ 20%) than the downward facing video. Cost-benefit optimization indicated that the forward facing video was more cost-effective at detecting broad scale change in coral, but less effective at detecting change in non-canopy forming algae compared to the downward video. Despite differences in detection of coral and algae, Deming regression analysis detected significant relationships between the percentages of common benthos recorded by both methods, enabling direct comparison between the data collected by each method at a broad scale level and for common benthos. Stereo-DOV surveys for fish can be considered a potential technique for the description of reef benthos that captures key elements of structural complexity, but is not necessarily suitable for taxonomic benthic monitoring.