Coral reefs have been more severely impacted by recent climate instability than any other ecosystem on Earth. Corals tolerate a narrow range of physical environmental stress, and increases in sea temperature of just 1 °C over several weeks can result in mass coral mortality, often exceeding 95% of individuals over hundreds of square kilometres. Even conservative climate models predict that mass coral bleaching events could occur annually by 2050. Unfortunately, managers of coral-reef resources have few options available to meet this challenge. Here, we investigate the role that fisheries conservation tools, including the designation of marine reserves, can play in altering future trajectories of Caribbean coral reefs. We use an individual-based model of the ecological dynamics to test the influence of spatially realistic regimes of disturbance on coral populations. Two major sources of disturbance, hurricanes and coral bleaching, are simulated in contrasting regions of the Caribbean: Belize, Bonaire, and the Bahamas. Simulations are extended to 2099 using the HadGEM1 climate model. We find that coral populations can maintain themselves under all levels of hurricane disturbance providing that grazing levels are high. Regional differences in hurricane frequency are found to cause strikingly different spatial patterns of reef health with greater patchiness occurring in Belize, which has less frequent disturbance, than the Bahamas. The addition of coral bleaching led to a much more homogenous reef state over the seascape. Moreover, in the presence of bleaching, all reefs exhibited a decline in health over time, though with substantial variation among regions. Although the protection of herbivores does not prevent reef degradation it does delay rates of coral loss even under the most severe thermal and hurricane regimes. Thus, we can estimate the degree to which local conservation can help buy time for reefs with values ranging between 18 years in the Bahamas and over 50 years in Bonaire, compared with heavily fished systems. Ultimately, we demonstrate that local conservation measures can benefit reef ecosystem services but that their impact will vary spatially and temporally. Recognizing where such management interventions will either help or fail is an important step towards both achieving sustainable use of coral-reef resources and maximizing resource management investments.