There has been rapid uptake of stereo-video-based sampling techniques to collect species, abundance and body-size information on fish assemblages and their associated habitats. Stereo-video methods provide highly accurate estimates of body-size, range and sampling area as well as a permanent record that can be cross-checked or resampled for additional data. Due to these advantages, diver-operated stereo-video (stereo-DOV) is becoming an increasingly popular alternative to underwater visual census. We provide a comprehensive guide for researchers using stereo-DOVs to survey fish assemblages and their associated habitat. Information on stereo-DOV design, video camera settings, field operations and video analysis are outlined. Stereo-DOV surveys permit rapid and simultaneous collection of data on fish diversity, abundance, length and behaviour as well as their associated habitat. However, biases associated with diver presence in the water and variation in detectability should be considered and are dependent on the location and focal species of the survey. We recommend using stereo-DOVs for diver-based surveys of non-cryptic fish assemblages, given the advantages described. An increased uptake of this methodology, following the standard procedures described herein, will reduce variation in methodology, assist in the synthesis of data on continental and global scales and provide accurate information to improve fisheries management and conservation.