Very little is known about the population dynamics of pine martens M artes martes and stone martens M artes foina in Europe, and even less about the impact of human activities, such as trapping, hunting and road traffic, on populations, despite their crucial importance for deriving reliable conservation and management strategies. We used multistate capture-recapture models to analyse capture-recapture histories that combined information on radio-tracking data and on recoveries of dead martens. This formulation, based on conditional probabilities, also allowed us to estimate the contribution of trapping and hunting to overall mortality, accounting for detection failures and trapping exposure. We applied this approach to estimate age- and sex-specific survivals of pine martens ( N = 40) and stone martens ( N = 38) in the same rural area in France, where both species could be legally trapped and hunted. Annual survivals of adults were similar, and estimated at 0.49 for both species [± 0.12 standard deviation ( sd) for pine martens; ± 0.10 sd for stone martens]. The annual probability of dying from trapping and hunting was estimated at 0.21 and 0.19 for adult pine and stone martens, respectively. At the individual level, a negative linear relationship was observed between stone marten's survival and trapping exposure, but this was not the case for pine martens, which experienced far lower exposure to trapping. These results confirm that trapping and hunting could have a large impact on marten populations. Vehicle collisions were the second major cause of mortality for both species in our study area. This paper presents the first estimates of survival for syntopic stone and pine martens in Europe and is an important step towards population dynamics modelling of these species. [ABSTRACT FROM AUTHOR]