Over the last twenty years, several institutes reconstructed past sea surface conditions, mainly focusing on the sea surface temperatures (SST) for more or less limited regions of the Atlantic Ocean during the Last Glacial Maximum (LGM). These reconstructions were based on different proxy data and employed different methods, too. Moreover, even the LGM time slice definitions were not alike in all cases, making it even more difficult to compile a consistent, ocean-wide set of temperatures that could be used for ocean modelling purposes. As a major improvement, the greater part of the research groups recently agreed on a unique definition of the LGM time slice and reevaluated their data. Although still based on different transfer functions and/or modern analog techniques, the new data are much more consistent than before. Under these auspices, we combine the available regional sets of Atlantic SST reconstructions and provide new gridded SST fields covering the complete Atlantic Ocean both for glacial February and August. For global modelling tasks, the new Atlantic data is incorporated into the world-wide CLIMAP reconstruction. From these sets and the available d18O data, we estimate the sea surface salinities (SSS) for both seasons. Monthly sets of SST and SSS are reconstructed following the PMIP guidelines by sine-interpolating the seasonal LGM-to-modern anomalies. Possible applications of these new data include: (1) validating coupled ocean-atmosphere models, (2) driving atmosphere-only models, (3) driving ocean-only models. With regard to the circulation and water masses of the Atlantic Ocean, we expect not only new insights into the glacial-interglacial changes of deep and intermediate water production in high latitudes but as well an improved understanding of still unresolved issues concerning the importance of upwelling strength and origins of intermediate waters for the biological production rates in the West-African upwelling areas.