Assessing the vulnerability of biological communities to anthropic pressures in marine systems may be challenging because of the difficulty to properly model each species' response to the pressure due to lack of information. One solution is to apply factor-mediated vulnerability assessment which combines (i) information on species ecological traits and conservation status organized in a matrix of so-called “vulnerability factors”, (ii) a conceptual model of how these factors affect species vulnerability, and (iii) data on the spatial distribution and abundance of each species issued from at-sea surveys. Such factor-mediated vulnerability assessment was originally introduced in the seabird–wind farm context by Garthe and Hüppop (2004. Scaling possible adverse effects of marine wind farms on seabirds: developing and applying a vulnerability index. Journal of Applied Ecology, 41: 724–734) and has since then been expanded to many case studies. However, the mathematical formulations that were proposed at that time are overly simplistic and may overlook critical components of the impact assessment. Our study briefly reviews the original approach and highlights its hidden assumptions and associated interpretation problems, for example, the overestimation of disturbance pressure to the detriment of collision, or the very high contribution of log abundances in vulnerability maps. Then, we propose a revised framework that solves these issues and permits easy transposition to other community-pressure case studies. To illustrate the usefulness and generality of the revised framework, we apply it to two case studies, one concerning the vulnerability assessment of a seabird community to offshore wind farms in the Bay of Biscay, and another focusing on the vulnerability assessment of the benthic megafauna community to trawling pressure in the Barents Sea.