Paleoecological indicators are examined as to their accuracy in reconstructing past biotic communities and environmental conditions, their utility in answering important questions about such communities and conditions, and the temporal and spatial scales over which they are effective. Next, environmental problems susceptible of paleoecological analysis are considered, as are the ecosystem and landscape properties that can be inferred from such an analysis. The usefulness of paleoecology in anticipating ecological "surprises" is then discussed. Finally, a set of conclusions and recommendations is presented.