Complex systems, such as the earth's climate, coral reefs, oceans, and social-economic systems, often react in a surprising way to gradual change. If conditions change gradually, the system may seem to respond little, until a critical tipping point is reached, after which the system may collapse completely. After such collapse, the original state of the system is notoriously hard to restore. Up till now, managers have found it very hard to predict such catastrophic transition without a thorough knowledge of the underlying mechanisms. Egbert van Nes and Marten Scheffer analyzed various models and came to the conclusion that there is a simpler way to predict a catastrophic transition. In their work, which appears in the June issue of The American Naturalist, they show that, after small disturbances, the system will recover much slower if a collapse is near. They argue that such slower recovery can serve as an early warning signal for upcoming shifts. In practice the recovery rate can be determined from small experiments or by analyzing the natural variations in a time series.