An overview of conditions associated with the Oklahoma-Kansas tornado outbreak of 3 May 1999 is presented, with emphasis on the evolution of environmental and supercellular characteristics most relevant to the prediction of violent tornado episodes. This examination provides a unique perspective of the event by combining analyses of remote observational data and numerical guidance with direct observations of the event in the field by forecasters and other observers. The 3 May 1999 outbreak included two prolific supercells which produced several violent tornadoes, with ambient parameters comparable to those of past tornado outbreaks in the southern and central Great Plains. However, not all aspects leading to the evening of 3 May unambiguously favored a major tornado outbreak. The problems that faced operational forecasters at the Storm Prediction Center are discussed in the context of this outbreak, including environmental shear and instability, subtle processes contributing to convective initiation, the roles of preexisting boundaries, and storm-relative flow. This examination reveals several specific aspects where conceptual models are deficient and/or additional research is warranted.