The desire to know the future is as old as humanity. For the tourism industry the demand for accurate foretelling of the future course of events is a task that consumes considerable energy and is of great significance to investors. This paper examines the issue of forecasting by comparing forecasts of inbound tourism made prior to the political and economic crises that engulfed Indonesia from 1997 onwards with actual arrival figures. The paper finds that current methods of forecasting are not able to cope with unexpected crises and other disasters and that alternative methods need to be examined including scenarios, political risk and application of chaos theory. The paper outlines a framework for classifying shocks according to a scale of severity, probability, type of event, level of certainty and suggested forecasting tools for each scale of shock.