Magellan has revealed an ensemble of impact craters on Venus that is unique in many important ways. We have compiled a data base describing the 842 craters on 89% of Venus' surface mapped through orbit 2578. (The craters range in diameter from 1.5 to 280 km.) We have studied the distribution, size-density, morphology, geology, and associated surface properties of these craters both in the aggregate and, for some craters, in greater detail. We find that (1) the spatial distribution of craters is highly uniform; (2) the size-density distribution of large craters (diameters ≥35 km) is similar to the young crater populations on other terrestrial planets but at a much lower density that indicates an average age of only about 0.5 Ga (based on the estimated population of Venus-crossing asteroids); (3) unlike the case on other planets, the density of small craters (diameters ≤35 km) declines rapidly with decreasing diameters because of atmospheric filtering; (4) the spectrum of crater modification differs greatly from that on other planets: 62% of all craters are pristine, only 4% are embayed by lavas, and the remainder are affected by tectonism, but none are severely and progressively depleted (as extrapolated from the size-density distribution of larger craters); (5) large craters have a progression of morphologies generally similar to those on other planets, but small craters are typically irregular or multiple rather than bowl shaped; (6) diffuse radar-bright or -dark features surround some craters, and 367 similar diffuse “splotches” with no central crater are observed; and (7) other crater features unique to Venus include radar-bright or -dark parabolic arcs opening westward and extensive outflows originating in crater ejecta. The first three of these observations are entirely unexpected. We interpret them as indicating that the planet's cratering record was erased by a global resurfacing event or events, the latest ending about 0.5 Ga, after which volcanic activity declined (but did not cease entirely). Since the last resurfacing event, a maximum of 10% of the planet has been resurfaced and only about 4% of the craters have been obliterated. Convective thermal evolution models support this interpretation (Arkani-Hamed and Toksoz, 1984). Observations 3–7 confirm quantitatively the expectation that the dense atmosphere of Venus has strongly affected the production of craters. Large impactors have been relatively unaffected, intermediate-sized ones have been fragmented and have produced overlapping or multiple craters, a narrow size range has produced shock-induced “splotches” but no craters, and the smallest bodies have had no observable effect on the surface. The number of craters eliminated by the “atmospheric filter” is enormous, about 98% of the craters between 2 and 35 km in diameter that Magellan might have observed on a hypothetical airless Venus. Unique crater-related features such as parabolas and outflow deposits demonstrate the roles of Venus' high atmospheric density and temperature in modifying the crater formation process. Finally, heavily fractured craters and lava-embayed craters are found to have higher than average densities along the major fracture belts and rifted uplands connecting Aphrodite Terra and Atla, Beta, Themis, and Phoebe regiones. These craters thus provide physical evidence for recent volcanic and tectonic activity at a low level.