In complex, job-shop-like production departments it is usually very difficult to predict the near-future logistic performance as well as to explain or diagnose objectively why and how a certain performance has been achieved. Presents a prediction and diagnosis method that has been developed and tested in two production departments. Describes how the method provides realistic logistic performance targets in the short term with respect to the throughput of a production department and order completion times. The method also quantitatively determines ex post the impact of occurred disturbances on the realized performance. In the pilot project the method provided a clearer insight into relationships between logistic key variables, gave decision support to the capacity allocation decision, and generated reliable performance targets for the short term. More importantly, the actual performance became more open to discussion due to the objective explanation of the achieved performance, which opens the way to performance improvements.