The gold coral Savalia savaglia (Cnidaria, Zoantharia) is a rare component of the mesophotic zone of the Mediterranean Sea and northeastern Atlantic Ocean. During two field campaigns along the Italian coast in the South Tyrrhenian Sea, two populations of this species were discovered. The specimens were filmed and photographed by means of a Remotely Operated Vehicle (ROV). To identify the role of local bathymetry and other derived variables on presence and distribution of S. savaglia we used a Habitat Suitability (HS) modeling technique based on Ecological Niche Factor Analysis (ENFA), utilizing high-resolution multibeam bathymetry and ROV data. Among the set of environmental variables derived from multibeam data, slope, rugosity, eastness and distance to rocks, appear to be the main variables involved in S. savaglia distribution, pointing out that the habitat differs considerably from the mean environmental conditions over the study area, and that S. savaglia ecological niche is significantly narrower than the available habitat. The HS map was developed to differentiate the sea floor into suitability classes. The comparison between suitability classes and presence data showed that the HS map is coherent with the observed spatial distribution of the species. The most suitable habitat for S. savaglia is characterized by a rough sea floor with rocks that is steeply sloped, oriented northeast, and within a water depth range of 34–77 m. Our study suggests that predictive modeling is an approach that can be applied to other deep coral species to locate areas with a suitable habitat. Considering the difficulties to reach the habitats in which these species live, this approach could be essential to planning further studies that help define areas where the species may be present.