Coral reefs provide significant evidence for former sea-level positions because of their geological preservation and suitability for dating. Interpretation of this evidence presumes an understanding of reef geomorphology, modern reef organism distributions, and environmental factors influencing them. Fossil reef terraces, formed during the last interglacial, marine oxygen isotope (MIS) substage 5e (~ 128–116 ka), are prevalent on many tropical shorelines and there has been ongoing debate as to the height reached by sea level during that highstand. Observations from numerous last interglacial sites suggest that sea level was at least 3 m above present sea level, implying less extensive icesheets than at present. An elevation of 6 m has commonly been adopted when correcting tectonically active sites for uplift. Recent compilations suggest elevations up to 8–9 m, but incorporate few observations from reefs where the last interglacial is found below sea level. Oscillation of sea level during MIS 5e has been interpreted from several sites, with recent studies inferring rapid rise of several metres at the end of the interglacial. These interpretations are at the limits to the precision with which corals can currently be dated and their palaeo-water depths inferred. It is not surprising that constraining last interglacial sea-level changes within uncertainties of less than 1–2 m remains controversial, considering sea-level variations recognised between reef sites in the Holocene, and observed geographical variation in isostatic or flexural adjustments. Fossil coral reefs on uplifting margins also provide clear evidence for MIS substages 5c and 5a, and those on Huon Peninsula indicate fluctuations related to Heinrich events (MIS 3). Interpretations show considerable variability between sites, with still greater uncertainties about sea-level timing and elevation during previous interglacials. Future study of extensive sequences of fossil reefs preserved on rapidly subsiding margins could address these uncertainties. Submerged reefs have already yielded important information about sea-level rise during the last deglaciation. Coring around Barbados and Tahiti, as well as on the Huon Peninsula, has produced a broadly consistent picture of ice melt, reflecting eustatic change since the last glacial maximum. These studies have shown the sensitivity of reefs to rapid sea-level rise associated with meltwater pulses, with some reefs drowning while others back-stepped. Integrated Ocean Drilling Program (IODP) expeditions to Tahiti, and recently the Great Barrier Reef, extended these records, but details of timing, nature and impact of deglacial meltwater pulses remain elusive. Studies of Holocene reefs have indicated different growth strategies; some kept up with sea level, while others caught up when sea level decelerated. Holocene sea level appears to have experienced a gradual rise up to present across the Caribbean, providing accommodation space for reefs to accrete vertically; whereas in the Indo-Pacific sea level has been near its present level since 7 ka, with many reef flats emergent following a slight fall of sea level caused by ocean siphoning. Microatolls on reef flats provide perhaps the clearest evidence of past sea-level position, but, in their absence, novel biological or other sea-level indicators are required to better constrain palaeo-water depths. There is an urgent need for further research from additional key reef locations, not only to decipher processes driving past sea-level change and its geographical variability, but also to better understand how coral reefs will respond in the context of future sea-level rise.