EU member states have to develop their Natura 2000 networks in their national waters to fulfill their conservation obligations regarding species and habitats listed in the Birds and Habitats directives. In France, a coastal network of Natura 2000 areas exists since 2008 but it had to be completed in offshore waters for some marine megafauna species. The SAMM aerial surveys (Aerial Census of Marine Megafauna) which occurred in winter 2011 and summer 2011–2012 over a large area comprising the whole metropolitan French Economic Exclusive Zone produced sighting data for species listed in the Birds and Habitats directives. These data produced different types of species distribution data: encounter rates and predicted densities by kriging and habitat modelling. Using these species distribution data, the aim of the present study was to compare these different types of inputs in the same conservation prioritization process to complete the existing Natura 2000 network in French waters. We ran prioritization analyses using the encounter rates only (scenario 1) then using the predicted densities provided by kriging and habitat modelling (scenario 2). We then compared the outputs of the two prioritization processes. The prioritization outputs were different but not in contradiction, with similar areas appearing as important to reach the conservation targets. Habitat models were thought to provide better pictures of seasonal species distributions and informed scientists about the phenology and ecology of species. However, the use of encounter rates as input data for the prioritization process in the Natura 2000 program is acceptable provided that sufficient survey effort is available.