Coral bleaching events are often associated with higher levels of coral mortality but when this occurs in the chronology of individual bleaching events is poorly documented. Knowing when mortality occurs is important for understanding molecular mechanisms and the putative adaptive significance of the response (the Adaptive Bleaching Hypothesis). In a detailed study of a coral bleaching event on the Great Barrier Reef, involving weekly and twice weekly repetitive observations of >200 individually marked corals over an 18 month period (∼16,000 observations), it is shown that bleaching in Acropora latistella, A. subulata and Turbinaria mesenterina was an acute, rapid response, occurring within days of a peak in seawater temperatures exceeding previously described thresholds. Subsurface light levels, measured over the duration of the event, were not anomalous. Full bleaching (i.e. whole colonies turning bone-white) and partial bleaching (white patches) was observed in the Acropora spp. whilst the T. mesenterina colonies typically paled to a light brown colour. Algal densities in bleached corals were 10–30% of those of normally pigmented corals (∼2.5 × 106 algae per cm2), and in this instance bleaching was clearly a sudden, isolated, stress event and not an extreme low-point in the seasonal fluctuation of the density of symbiotic algae. Bleached corals were associated with high levels of partial and whole-colony mortality, but mortality was exclusively limited to the two Acropora spp. Importantly, most of this mortality was recorded in surveys conducted 1 and 2 weeks after bleaching was first observed, and for A. latistella as little as 1 week after bleaching was first observed. This suggests that in this particular bleaching event, for the Acropora species, that bleaching and mortality were intimately linked: this in turn suggests it was a pathological phenomenon. The study highlights a problem in the adaptive bleaching hypothesis, whereby significant levels of mortality can occur in a bleaching event before any chance for subsequent recombination of the host-symbiont unit. It is argued that in order to further evaluate the significance of bleaching as a potentially adaptive mechanism, bleaching-induced and bleaching-related mortality have to be fully considered. It is necessary to incorporate the cost (in terms of mortality) of a bleaching event, the recurring cost of reverting to the original, mortal, stress–prone combination after the event, and the higher cost associated with forming a maladaptive combination.