Catastrophic events such as oil spills, hypoxia, disease, and major predation events occur in marine ecosystems and affect fish populations. Previous evaluations of the performance of spatial management alternatives have not considered catastrophic events. We investigate the effects of local and global catastrophic events on populations managed with and without no-take marine reserves and with fishing mortality rates that are optimized accounting for reserves. A spatial population dynamics model is used to explore effects of large, catastrophic natural mortality events. The effects of the spatial spread, magnitude, probability of catastrophe, and persistence of a catastrophic event through time are explored. Catastrophic events affecting large spatial areas and those that persist through time have the greatest effects on population dynamics because they affect natural mortality nonlinearly, whereas the probability and magnitude of catastrophic events result in only linear increases in natural mortality. The probability of falling below 10% or 20% of unfished abundance was greatest when a no-take marine reserve was implemented with no additional fishing regulations and least when a no-take marine reserve was implemented in addition to the maintenance of optimal fishing mortality in fished areas. In the absence of implementation error, maintaining abundance across space using restrictions on fishing mortality rates, regardless of the existence of a no-take marine reserve, decreased the probability of falling below 10% or 20% of unfished abundance.