The threat of an abrupt and extreme rise in sea level is widely discussed in the media, but little understood in practise, especially the likely impacts of such a rise including a potential adaptation response. This paper explores for the first time the global impacts of extreme sea-level rise, triggered by a hypothetical collapse of the West Antarctic Ice Sheet (WAIS). As the potential contributions remain uncertain, a wide range of scenarios are explored: WAIS contributions to sea-level rise of between 0.5 and 5 m/century. Together with other business-as-usual sea-level contributions, in the worst case this gives an approximately 6-m rise of global-mean sea level from 2030 to 2130. Global exposure to extreme sea-level rise is significant: it is estimated that roughly 400 million people (or about 8% of global population) are threatened by a 5-m rise in sea level, just based on 1995 data. The coastal module within the Climate Framework for Uncertainty, Negotiation and Distribution (FUND) model is tuned with global data on coastal zone characteristics concerning population, land areas and land use, and then used for impact analysis under the extreme sea-level rise scenarios. The model considers the interaction of (dry)land loss, wetland loss, protection costs and human displacement, assuming perfect adaptation based on cost-benefit analysis. Unlike earlier analyses, response costs are represented in a non-linear manner, including a sensitivity analysis based on response costs. It is found that much of the world's coast would be abandoned given these extreme scenarios, although according to the global model, significant lengths of the world's coast are worth defending even in the most extreme case. This suggests that actual population displacement would be a small fraction of the potential population displacement, and is consistent with the present distribution of coastal population, which is heavily concentrated in specific areas. Hence, a partial defence can protect most of the world's coastal population. However, protection costs rise substantially diverting large amounts of investment from other sectors, and large areas of (dry)land and coastal wetlands are still predicted to be lost. Detailed case studies of the WAIS collapse in the Netherlands, Thames Estuary and the Rhone delta suggest greater abandonment than shown by the global model, probably because the model assumes perfect implementation of coastal protection and does not account for negative feedbacks when implementation is imperfect. The significant impacts found in the global model together with the potential for greater impacts as found in the detailed case studies shows that the response to abrupt sea-level rise is worthy of further research. © 2008 Springer Science+Business Media B.V.