Forecasting methods, as a control tool, aim to predict final project outcomes and play an important role in assessing the performance of construction projects. In construction, accuracy has been the most important criterion used to assess forecasting performance. However, in addition to accuracy, forecasting systems must provide timely information about project status so corrective action can be taken when required. Therefore, researchers have introduced timeliness as an additional forecasting performance measure. A preliminary study found that some forecasting methods with good timeliness indices do not necessary predict deviations with precision. Furthermore, the cost predictability indices proposed in the literature do not provide sufficient information upon these two desirable features of forecasting methods (timeliness and accuracy). Currently, researchers and practitioners are looking for appropriate methodologies that allow project managers to assess forecasting performance. This study assessed the performance of construction project cost forecasting methods using data envelopment analysis (DEA), based on timeliness and accuracy indices. Preliminary results showed that there is statistical evidence to indicate that the ranking obtained using DEA differs from that obtained using the traditional forecasting performance indices.