The effectiveness of a remotely operated vehicle (ROV) to describe the fish communities of three gas platforms located offshore Crotone (Italy, Ionian Sea) was investigated by comparing its observations with underwater visual censuses (UVCs). The study was carried out at two depth layers (0-6 and 12-18 m). Moreover, the ROV was used to survey three deeper depth layers up to 76 m. Overall, the ROV surveys failed to give a truthful representation of the fish communities underestimating the number of species and their abundances as compared to UVCs. The main discrepancies in data regarded cryptobenthic and nekto-benthic species, whereas the ROV proved to be a suitable method to census low-mobile and abundant planktivorous species. The differences between the fish assemblage described by the ROV, with respect to the one depicted by UVC, should be considered in the light of the technical limits of the recording camera, whose resolution and field of vision is clearly lower than the diver's eye. In addition, video images did not allow for the acquisition of a correct estimate of the distance between the individuals and the platform structures. This led, almost certainly, to an under-or over-estimation of fish abundance as regards to the censused volume. In spite of this, as a result of its capacity to reach depths inaccessible to scuba divers and then to add complementary information, the ROV could be used jointly with UVCs, in studies having as their objective the description of the fish communities associated with offshore platforms.