This work points out the ability of the electronic nose to detect the presence of unburned diesel in engine oil, in comparison with standard methods based on distillation or gas chromatography. The system is based on an array of different gas micro-sensors whose sensitive layers are metal oxide thin films deposited by sol-gel technique on Si substrates. The sensor array, exposed to the volatile chemical species of different engine oil samples contaminated by diesel fuel, resulted to be appreciable sensitive to detect the fuel contamination. The results of Principal Component Analysis, used as exploratory data analysis technique, and the success rate of the classification, based on Gaussian mixture model, prove the good performance of the system in discriminating different diesel fuel diluted lubricating oil samples.