A finite element model for ice sheets has been modified to include an accounting of basal melt water. The calculated amount of basal water is used to find the velocity of ice sliding over the bed. The interplay of the basal water and sliding velocity is investigated by a series of modeling experiments. Simulations of the interception, growth, and de-glaciation of the Northern Hemisphere are considered. The simulations are driven with a climate controlled by the GRIP ice core data. The experiments contrast a number of different mechanisms for relating the amount of basal water and the sliding velocity. The mechanisms range from no sliding whatsoever, to a scheme in which the sliding velocity is proportional to the amount of basal water. The modeled results demonstrate that choice of sliding mechanism has a dramatic effect on the glaciation cycle. Furthermore, modeling results indicate that the sliding velocity must be related to the basal water depth to correctly simulate a complete glacial cycle. The relation is important to inception and all other phases of a glacial cycle. The volume of the ice at the last glacial maximum can be compared to the estimated sea level rise to further constrain the new modeling approaches.