Warm-water growth and survival of corals are constrained by a set of environmental conditions such as temperature, light, nutrient levels and salinity. Water temperatures of 1 to 2°C above the usual summer maximum can trigger a phenomenon known as coral bleaching, whereby disruption of the symbiosis between coral and dinoflagellate micro-algae, living within the coral tissue, reveals the white skeleton of coral. Anomalously cold water can also lead to coral bleaching but has been the subject of limited research. Although cold-water bleaching events are less common, they can produce similar impacts on coral reefs as warm-water events. In this study, we explored the effect of temperature and light on the likelihood of cold-water coral bleaching from 1998-2017 using available bleaching observations from the Eastern Tropical Pacific and the Florida Keys. Using satellite-derived sea surface temperature, photosynthetically available radiation and light attenuation data, cold temperature and light exposure metrics were developed and then tested against the bleaching observations using logistic regression. The results show that cold-water bleaching can be best predicted with an accumulated cold-temperature metric, i.e. ‘degree cooling weeks’, analogous to the heat stress metric ‘degree heating weeks’, with high accuracy (90%) and fewer Type I and Type II errors in comparison with other models. Although light, when also considered, improved prediction accuracy, we found that the most reliable framework for cold-water bleaching prediction may be based solely on cold-temperature exposure.