This paper gives the yearly (85)Kr emissions of all known reprocessing facilities, which are the main sources of (85)Kr in the atmosphere since 1945, for the years 1945 until 2000. According to this inventory 10,600 PBq (Peta=10(15)) of (85)Kr have been globally emitted from the year 1945 until the end of 2000. The global atmospheric inventory at the end of the year 2000 amounts to 4800 PBq. These emissions have been incorporated into the ECHAM4 atmospheric general circulation model as point sources. Monthly mean model results are compared with measurements made at different locations and times. The influence of each source on the measured concentrations at various locations is studied. The calculated concentrations are found to give reasonably good agreement with the observations, indicating that the emission inventory is realistic. Although, at all northern hemispheric observation sites the model tends to slightly overestimate the concentrations. A possible reason for this overestimation can be found in model features (coarse resolution in time and space). The most prominent discrepancy that is consistently repeated at all northern hemispheric stations occurs in the early 1990s. This could most likely be related to an overestimate of sources. Possibly, the Russian emissions declined earlier than assumed in the current database. Another discrepancy between observations and simulations indicating an incompleteness of the release data is found at some southern hemispheric sites. The variability of their observations could only be explained by regional sources. However, several spikes occur after 1992 when no reprocessing facility is known to be in operation in the southern hemisphere. Production of isotopes for radiopharmaceuticals like technetium-99m from highly enriched uranium is the most likely explanation.