Capabilities of remotely operated vehicles (ROVs) have increased substantially in the last decade, and mini-ROV designs are now able to conduct visual research frequently conducted by snorkellers or divers in shallow marine environments. There are logistical, financial and experimental benefits of using mini-ROVs over snorkellers or divers, yet the adoption of mini-ROVs for common shallow underwater research tasks has not been widespread. To assess the capabilities of mini-ROVs to sample fish communities we compared the results produced by a mini-ROV to that of snorkellers for performing two of the most common marine video-based research activities (1) underwater visual fish census and (2) observing and tracking fish behaviour. Results of both activities suggested that the fish community observed by the mini-ROV was not distinguishable to that observed by the snorkellers, however, the mini-ROV detected significantly more fish (39% higher abundance) and greater diversity (24% higher). When tracking butterflyfish behaviour, video obtained from the mini-mini-ROV was as efficient as a snorkeller at finding and tracking individuals. Video from the mini-ROV produced comparable responses to that from snorkellers with hand-held GoPros, although over the course of tracks the response between the two methods differed, with a decrease in refuge time for snorkeller video and an increase in tailbeat rate for the mini-ROV video. Our study shows that video obtained from mini-ROVs can be used for research in shallow marine environments when direct manipulations are not required. We predict the research capabilities of mini-ROVs to increase substantially in the coming years, which should cement the use of this tool for research across all marine environments.