Two years after California's devastating Loma Prieta earthquake of October 17, 1989, some surprising results are emerging from the study of the faulting that created the earthquake. These surprises include a significant discrepancy between the two most reliable methods of locating the fault break, the very short duration of sliding on the fault in the earthquake, and the discovery of a magnitude 5 foreshock at the beginning of the event. The Loma Prieta earthquake marks an important watershed in the study of the earthquake source. Never before have such complete measurements been obtained for both the distortion of the Earth's surface and the strong shaking produced during one earthquake. As a result, detailed comparisons between models of the earthquake developed from independent observations permit geophysicists to test some of their fundamental assumptions about the nature of earthquakes and to evaluate the ability of different methods to decipher the earthquake.