The 28 September 2004 M 6.0 Parkfield earthquake, a long-anticipated event on the San Andreas fault, is the world's best recorded earthquake to date, with state-of-the-art data obtained from geologic, geodetic, seismic, magnetic, and electrical field networks. This has allowed the preearthquake and postearthquake states of the San Andreas fault in this region to be analyzed in detail. Analyses of these data provide views into the San Andreas fault that show a complex geologic history, fault geometry, rheology, and response of the nearby region to the earthquake-induced ground movement. Although aspects of San Andreas fault zone behavior in the Parkfield region can be modeled simply over geological time frames, the Parkfield Earthquake Prediction Experiment and the 2004 Parkfield earthquake indicate that predicting the fine details of future earthquakes is still a challenge. Instead of a deterministic approach, forecasting future damaging behavior, such as that caused by strong ground motions, will likely continue to require probabilistic methods. However, the Parkfield Earthquake Prediction Experiment and the 2004 Parkfield earthquake have provided ample data to understand most of what did occur in 2004, culminating in significant scientific advances.