Effective conservation requires monitoring and pro-active risk assessments. We studied the effects of at-sea mortality events (ASMEs) in marine mammals over two decades (1990-2012) and built a risk- based indicator for the European Union’s Marine Strategy Framework Directive (MSFD). Strandings of harbour porpoises (Phocoena phocoena), short-beaked common dolphins (Delphinus delphis), and striped dolphins (Stenella coeruleoalba) along French coastlines were analysed using Extreme Value Theory (EVT). EVT operationalises what is an extreme ASME, and allows the probabilistic forecasting of the expected maximum number of dead animals assuming constant pressures. For the period 2013-2018, we forecast the strandings of 80 harbour porpoises, 860 common dolphins, and 57 striped dolphins in extreme ASMEs. Comparison of these forecasts with observed strandings informs whether pressures are increasing, decreasing or stable. Applying probabilistic methods to stranding data facilitates the building of risk-based indicators, required under the MSFD, to monitor the effect of pressures on marine mammals.