In coastal regions, several hazards may lead to floods, and if they occur concurrently, the damage will be higher than for the hazards individually. The paper outlines an approach for carrying out a risk analysis with several hazards and applies it on a case study in Greater Copenhagen where two hazards, rainfall and sea surge, are both important. The core in the methodology is the application of copula functions as an extension of one-dimensional risk analysis and projections of future climatic changes. The results for Greater Copenhagen indicate that the dependence between the hazards is weak and that climate change most likely will not increase the correlation. The overall change in flood return periods over a forecast horizon of 110 years are estimated to decrease by one to three orders of magnitude.