The mean Arctic January climatology and its interannual variation have been examined by simulations with a regional climate model of the Arctic atmosphere. To this end, an ensemble of monthlong simulations (January of the 11 years 1985-1995) has been investigated in which the ensemble is large enough to represent a broad range of climatic conditions. The model produces credible simulations of the meteorological patterns. Only small deviations occur between simulations and European Centre for Medium-Range Weather Forecasts (ECMWF) analyses; that is, over most of the model area these differences in sea level pressure, 850 hPa temperature, and 500 hPa height are below 3 hPa, 1 K, and 5 m, respectively. Larger deviations (up to 5 hPa in the sea level pressure and 20 m in the 500 hPa height) are found over parts of the Arctic Ocean, which seems to be related to the crude sea ice representation at the lower model boundary and deficiencies in the planetary boundary layer parameterization. It is shown that the dynamical aspects of the interannual variability can be adequately captured by the model simulations; that is, the maximum of the model bias of the dynamical variables is significantly smaller than the interannual variability throughout the entire domain. To gain more insight into the spatial and temporal structures of the model's variability, an empirical orthogonal function (EOF) analysis has been applied to determine the most significant structures in the fluctuations of the monthly mean dynamical fields. EOF 1 of the 500 hPa height field describes a regime with well-pronounced polar vortex and corresponds to the "Arctic Oscillation", whereas EOFs 2 and 3 show wave structures. A pronounced interannual variability is noticed in the time series of the amplitudes of the EOFs. Copyright 1999 by the American Geophysical Union.