The representation of the atmospheric hydrologic cycle over the Arctic in version 3.0 of NCAR's Community Climate System Model has been compared to existing observations and previous studies. In addition to examining mean fields and annual cycles, composite analysis has been used to identify atmospheric circulation patterns associated with hydrologic variability in the model. Good agreement with observations is found in the North Atlantic and the Ob, Yenisey and Lena river watersheds in Eurasia, where annual cycles are reasonable and the model appears to be reproducing realistic precipitation events. However, the model's hydrology is flawed within the Mackenzie River Basin (MRB), which recieves excess cold season precipitation. Composites of modeled sea level pressure and 500 mb geopotential height show patterns similar to those found during lee-side cyclogenesis in the southern end of the MRB, suggesting the model is producing too many of these cyclones. Keeping these strengths and weaknesses in mind, the adjustment of the Arctic hydrologic cycle to climate change scenario runs is examined, with particular attention paid to the well-modeled Eurasian watersheds and Atlantic sector.