Over the last four years, a heated debate has erupted over the validity and utility of the seismic gap model. Kagan and Jackson [1991] have demonstrated that the seismic gap hypothesis as formulated by McCann et al. [1979] and Kelleher et al. [1973] (that earthquakes are more likely to occur in areas that have not recently had large earthquakes) has failed to predict where major earthquakes are more likely to occur. Kagan and Jackson [1991] interpret this failure to mean that earthquakes are clustered in time, adding that because plate tectonics only provides long-term constraints on slip rates, it does not imply the regularity of earthquakes. Another possible interpretation of the failure of the seismic gap hypothesis as proposed by McCann et al. [1979] is that it is too crude to predict major earthquakes reliably. A more rigorous statement and test of the seismic gap hypothesis needs to be made. It would require the determination of the slip distribution in seismic zonesPand hence detailed seismic source modelingPand would test whether or not future earthquakes are most likely to occur in areas that have not slipped recently. It is important that we understand the nature of earthquake recurrence and interaction [ Harris and Simpson, 1992; Stein et al., 1992; Jaumé and Sykes, 1992; Beroza and Zoback, 1993; Hill et al., 1993] because our ability to update seismic hazard after major earthquakes occur depends on it. Seismic source modeling will play a key role in this effort.