The 1990s has seen the proposal of several lunar satellite missions, some of which would measure gravity and topography, but none of which have yet been approved. Planning for these missions led to the observation that the available lunar gravity models predicted vastly different orbit behavior, resulting in large uncertainties in the predicted fuel requirements. This in turn led to a complete reanalysis of the Doppler tracking data from Lunar Orbiters I, II, III, IV, and V and the Apollo 15 and 16 subsatellites by Konopliv et al. [1993a]. Their 60 x 60 model was developed using the priority constraints discussed earlier, which is necessary to constrain the gravity model on the farside of the Moon. This gravity model predicts similar orbit behavior to the Bills and Ferrari [1980] model, and it is now believed that other models predicting different behavior were in error. The improvements gained by Konopliv et al. [1993a] were derived almost entirely through improved data processing and solution techniques, since the data employed are the same as for earlier models. The other major development in modeling the lunar gravity field has been the success of the Clementine mission. Zuber et al. [1994] obtained significant improvements to the gravity model through the analysis of the Clementine tracking data, although the gravity model on the lunar farside is still not well determined. In addition, the Clementine altimeter has provided substantial improvements in the topography model for the Moon [ Zuber et al., 1994], which has allowed unprecedented geophysical studies of the Moon's crustal strength and interior.