In this study we present the first results of a new isotope-enabled general circulation model set-up. The model consists of the fully coupled ECHAM5/MPI-OM atmosphere-ocean model, enhanced by the JSBACH interactive land surface scheme and an explicit hydrological discharge scheme to close the global water budget. Stable water isotopes (H2O)-O-18 and HDO have been incorporated into all relevant model components. Results of two equilibrium simulations under pre-industrial and Last Glacial Maximum conditions are analysed and compared to observational data and paleoclimate records for evaluating the model's performance in simulating spatial and temporal variations in the isotopic composition of the Earth's water cycle. For the pre-industrial climate, many aspects of the simulation results of meteoric waters are in good to very good agreement with both observations and earlier atmosphere-only simulations. The model is capable of adequately simulating the large spread in the isotopic composition of precipitation between low and high latitudes. A comparison to available ocean data also shows a good model-data agreement; however, a strong bias of overly depleted ocean surface waters is detected for the Arctic region. Simulation results under Last Glacial Maximum boundary conditions also fit to the wealth of available isotope records from polar ice cores, speleothems, as well as marine calcite data. Data-model evaluation of the isotopic composition in precipitation reveals a good match of the model results and indicates that the temporal glacial-interglacial isotope-temperature relation was substantially lower than the present spatial gradient for most mid-to high-latitudinal regions. As compared to older atmosphere-only simulations, a remarkable improvement is achieved for the modelling of the deuterium excess signal in Antarctic ice cores. Our simulation results indicate that cool sub-tropical and mid-latitudinal sea surface temperatures are key for this progress. A recently discussed revised interpretation of the deuterium excess record of Antarctic ice cores in terms of marine relative humidity changes on glacial-interglacial timescales is not supported by our model results.