In this paper, the distillation curve points and certain cold properties of diesel fuel fractions were determined by means of mid infrared spectroscopy (IR) and artificial neural networks (ANNs). Distinct ANN models were developed for the prediction of the distillation curve with 5% recovery increment, and for the pour (PP) and cloud (CP) points. The properties of interest were measured experimentally following standard ASTM methods. The developed prediction algorithms were tested on unseen data and exhibited high level of accuracy, comparable to the one of the standard laboratory methods. The presented algorithms can be used as a rapid and reliable tool in quality control applications of the various fractions involved in diesel fuel production, since they can determine several important properties simultaneously based on a single and fast IR measurement. (c) 2005 Elsevier Ltd. All rights reserved.