Several features of the Viking Mars atmospheric water detection observations have suggested that the residual north polar cap on Mars is a major source of atmospheric water during northern summer and that the water it supplies to the atmosphere is transported equatorward to low northern latitudes by summer's end. We have explored the role of the north residual cap in the current Martian water cycle by constructing models to assess the ability of the cap to supply water to the atmosphere and the ability of the atmosphere to transport it out of the polar regions. From our sublimation model we conclude that the residual cap can sublime the 7¿1014 g of water that appears in the atmosphere after Ls=80¿ provided surface winds are high (~10 m/s), the cap emissivity is low (~0.9), or the exposed soil in polar regions contains ice that is accessible to the atmosphere. In our view these requirments seem unlikely, and we are drawn to the conclusion that other sources of water exist. This conclusion is further supported by the results of two-dimensional transport simulations which show that most of the water sublimed from the cap remains at high latitudes. The polar circulation simply lacks the intensity and scale to move water very far from the cap. A weak circulation places additional constraints on cap sublimation rates since abundant clouds and significant ice deposits would form just equatorward of the cap edge, according to the model, if the cap is allowed to sublime 7¿1014 g of water. Such predictions are inconsistent with available data. We suggest that the nature of the additional source is water desorbing from the nonpolar regolith. ¿ American Geophysical Union 1990