A new major frontier of geological research, which was initiated in the 1970's, involves predicting future geologic trends or events through study of the present and past, rather than trying to understand the past, often using what one knows about the present. Like most scientific frontiers, this one began from practical considerations-environmental concerns. The lack of formal recognition of this frontier results from fragmentation among many Federal agencies and highly focused mission-oriented programs (e.g., earthquake prediction, CO2, nuclear-energy safety, etc.). Most programs aim to predict only the next 50-100 years, but much longer periods of the past need to be studied to do this. Nuclear-waste disposal has sometimes been considered in terms of the next million years, a period of time permitting significant and broad geologic changes. Decreasing public interest in environmental concerns relegates many questions from the realm of applied research back to that of basic research. Most of these questions are so fascinating, however, that the frontier is still worth pursuing. Such questions include whether a phenomenon will or will not take place and the rates at which it can develop (e.g., how fast do rifts form, how fast can a caldera event begin, and how quickly can a glacial maximum arrive?). Common elements of all studies include the historic record, trends in the Quaternary, analogues in various periods of the geologic time scale, and allowance for phenomena never experienced before. Other examples of studies include the Cretaceous as a period of a climatic extreme, an especially interesting time period; establishing the amount of paleocloudiness, a particularly challenging and important research area; acid rain as a possible new phenomenon. Geochemistry has much to contribute to this frontier science. ?? 1983.