The laboratory, where variables can be measured and controlled, is perhaps the most efficient place to test scientific theory, yet in general empirical financial research neglects experimental finance results. We link laboratory findings to actual, or field, data, applying the Smith, Suchanek and Williams  experimental market model to Australian stock exchange data. We introduce modifications that improve the model's fit to field market conditions. The experimental model is a reliable predictor of field market bubble bursts in more than 50% of the cases we test, and our modifications improve the performance to 77% of the cases. Our results suggest that experimental financial market results should be accorded more attention in empirical research.