The IPCC Fourth Assessment Report (AR4) stated that global sea level rose by about 120 m between the Last Glacial Maximum (LGM) and 3–2 ka and did not rise again until the late 19th century. A review of relevant literature demonstrates that the rise may have been larger (by 10–15 m), that the timing of stabilisation is poorly constrained (within the last 7 ka BP), and that the onset of modern rates of sea-level rise is not synchronous and uniform across the globe. The LGM lowstand as defined in the AR4 is based on the ‘classic’ Barbados sea-level history, but the sea-level curve from this island may not be an accurate ‘eustatic’ approximation if glacial isostatic adjustment (GIA) processes were significant, as suggested by various authors. The Barbados sea-level curve revealed the existence of meltwater pulse (MWP) 1a at around 14 000 cal. a BP. Questions about the source and causal mechanism of MWP 1a are still being explored by field scientists and GIA modellers. These problems have direct relevance for contemporary measurements of ice mass changes, because space geodetic measurements rely on information provided by GIA models to separate land motion from ice mass changes. Global patterns of Holocene sea-level change are controlled by regional sea-level variability, not only through isostatic and tectonic processes, but also through steric effects, longer-wavelength gravitational changes produced by changing ice–ocean mass flux and hemispheric-scale perturbations in the Earth's rotation. ‘Eustasy’ is therefore merely a concept, not a measurable quantity. Constraining the late Holocene ice melt contribution to global sea-level rise and determining regional sea-level rise patterns in the 18th and 19th centuries are important future tasks for sea-level scientists; these will establish the baseline contributions to 20th- and 21st-century sea-level rise and will help to determine the causes of modern rates of sea-level rise. Constraining historical sea-level accelerations and short-term ice-mass flux will aid models of ice-sheet dynamics and help to reduce current uncertainties in future sea-level predictions. The AR4 shows that regional sea-level variability is clearly present in sea-level change patterns for the past 50 a and an important challenge for the next IPCC assessment is to incorporate this variability into regional future sea-level rise scenarios. Copyright © 2009 John Wiley & Sons, Ltd.