Model-based control has had a tremendous impact in the process industries. If the process changes significantly, the dynamic model may no longer be adequate and thus the model-based control scheme may result in poor performance. However, poor agreement between model predictions and output data does not necessarily imply that model re-identification is required. This paper addresses the important issue of deciding when re-identification of the process model is required. Using metrics based on principal component analysis and the Akaike information criterion, it is shown how a relatively short, closed-loop test can be used to screen for changes in the process parameters. The utility of the approach is demonstrated by two simulation examples.