It is the thesis of this Article that the ultimate success of the National Environmental Policy Act (NEPA) is strongly dependent upon reasonably complete and timely scientific information. Implementation of NEPA requires scientific descriptions of environmental quality and productivity, and the essence of the environmental impact statement (EIS) is a capability to anticipate the environmental consequences of a proposed action. Prediction of ecosystem behavior includes both degradation from technological change and recovery or improvement due to pollution abatement.

The state of knowledge in environmental sciences fundamental to NEPA is unsatisfactory compared to the magnitude of the management decisions being made and the values at stake. The uncertainties of descriptions and predictions in ecosystem analysis are greater than in other natural sciences but the expectations of legislators and administrators for complete, unambiguous, and verified information remain high. Some strategies are developing for optimizing decision making in the face of uncertainty and these can be useful in implementing NEPA.