Natural mortality rate is a key parameter in ecology and fisheries, but it may be difficult to estimate for methodological reasons and because it is highly variable and often confounded with other factors such as fishing mortality and migration. This is especially true for crustaceans in general, because age determination is problematic, and for snow crab (Chionoecetes opilio) in particular because density-dependent processes may lead to highly variable mortality rates across life history stages. In this context, we developed an original method to estimate the natural mortality rate of adult (i.e. terminally molted) female snow crab, and its consequences on egg production. This new method relies on shell condition, carapace width and abundance of adult females and was applied to a time series (1991–2012) of annual trawl survey data for a snow crab population in the Gulf of St. Lawrence, eastern Canada. Two natural mortality estimates were provided by the method depending on two distinct assumptions about the survey. Both estimates (0.66 and 0.78 year–1) were high compared to previous estimates for snow crab. These values imply that female life expectancy after terminal molt was short and that primiparous females (first-time spawners) contributed a large share (at least 81%) of the total number of eggs produced by the case-study population over the period 1992–2010.