Studies of marine reserves typically focus upon differences in the size and abundance of target organisms inside versus outside reserve borders, but they seldom provide spatially explicit measurements of how reserves influence mortality rates. This study investigated mortality rates for female California spiny lobster (Panulirus interruptus) at multiple sites inside and outside of three marine reserves at the Santa Barbara Channel Islands, California, USA. Mean total mortality (Z) of female lobsters was lower at sites inside reserves (Z = 0.22 [± 0.05 SE]) than at sites outside reserves (Z = 0.59 [± 0.02 SE]). Mean mortality at all sites inside reserves, and among sites near reserve centres (where Z = 0.17 [± 0.05 SE]), was similar to estimates of natural mortality for other temperate spiny lobster species. Among sites inside reserves, there was a positive relationship between mortality and proximity to reserve borders, but this relationship was absent among sites outside reserves. Mortality estimates were much more variable among sites inside reserves than at sites in fished areas. This variation is probably due to differential emigration rates from the three reserves, as well as site-specific ecological factors that influence population structure, demonstrating the importance of spatially explicit reserve sampling and understanding how ecological heterogeneity influences fisheries models.