Despite recent advances there is still debate regarding the volume and areal extent of ice sheets in the Northern Hemisphere since the Last Glacial Maximum. Ice sheet models are uniquely placed to provide glaciologically-realistic insights into the problem. However there are unknown parameters in ice sheet models whose sensitivity must be determined before the results of ice sheet models can be accepted with confidence.To examine these sensitivities we simulate the retreat of the Northern Hemisphere ice sheets since the LGM using the Huybrechts thermo-mechanically coupled ice sheet model. A time dependent forcing is used, consisting of modeled LGM temperature and precipitation time slices from the UKMO Paleo-GCM (PMIP) model. Modeled surface temperature and precipitation differences between LGM and present day are used as inputs to the ice sheet model. The GRIP ice core record is rescaled to equal zero at present day and unity at LGM, to represent a time dependent 'glacial index'. The UKMO fields are then rescaled by this glacial index to generate time dependent changes in ice volume over an entire glacial cycle.With this forcing, the Huybrechts' ice sheet model generates a history of ice sheet extent very similar to that suggested by glacial-geomorphologic data. We therefore adopt this model as a baseline 'reference model' with which to compare results generated when changing ice sheet model parameters. Specific parameters examined are those related to flow enhancement, basal sliding, thermal properties, glacial isostasy and mass balance. The parameters which are found to most significantly vary ice sheet volume over a glacial cycle are flow enhancement and mass balance. We conclude that when a more sophisticated atmospheric forcing is used with an ice sheet model, it is the sensitivity of these parameters which must be most fully examined.