Monitoring species’ response in marine protected areas is important for informing both the management of those areas and the establishment of additional protected areas. Populations of spiny lobsters Jasus edwardsii were monitored in eight New Zealand marine reserves for up to 34 years. The populations displayed highly variable responses to protection. While a few showed rapid (within 1–2 years of protection) increases in abundance, others showed little response even after a decade of protection. Some reserves displayed little initial recovery, then a sudden increase following several years of protection, while others displayed significant declines in abundance following initial recovery. Marine reserves located in areas with initially high densities of juveniles tended to have rapid recovery, but aspects of reserve design had no significant influence on the recovery rate. Variability among recovery trajectories also suggests that supply-side dynamics may be a key driver of lobster recovery. Densities of legal-sized lobsters were positively correlated with reserve age, but the abundance of juvenile lobsters increased in all but one reserve, indicating enhanced recruitment, survival and/or movement of juvenile lobsters into reserves. It is important to consider the placement of reserves, with respect to potential levels of larval supply, when establishing marine reserves for either conservation or fisheries management purposes and for evaluating their effectiveness.