Abstract
This study compares Italy's preliminary industrial production index with the final data and with the performance of alternative forecasts. It also presents a methodology for providing more accurate estimates of the index than the original official data. The improvement in accuracy and bias is obtained using the information contained in the two time series themselves.