Crater production and obliteration are modeled for the plains of Venus, using: (1 ) the observed distribution of Venus-crossing asteroids and comets, (2) viscous relaxation of crater topography, and (3) erosion and burial by atmospheric, volcanic, and tectonic processes. Crater lifetimes are assumed to be proportional to crater depths for both classes of obliterative processes although the individual criteria vary. An average crater retention age between 0.4 to 2.0 Ga is estimated for plains under the assumption that craters are produced and not removed. The range is driven by uncertainty in identifying degraded impact as opposed to volcanic craters. On the other hand, crater retention ages greater than about 1.6 Ga are unlikely if viscous relaxation operates without loading of crater floors by burial. Our preferred model has plains subject to crater production and obliteration processes that vary over both space and time. In some areas, radar-bright crater ejecta haloes are preserved for long periods of time because volcanism, tectonism, and weathering occurs at rates « 1 km/Ga. Viscous relaxation has probably generated numerous shallow craters in these relatively quiescent regions. In other areas, volcanism and tectonism have resurfaced the terrain at rates greater than several km/Ga. The global coverage and high resolution SAR and altimetry data expected from Magellan will allow testing of this model, based on detailed crater observations (diameter, depth distributions; morphologic criteria; surface scattering properties) and their association with volcanic and tectonic features.