Intelligent lubricating oil analysis is a procedure in condition-based maintenance (CBM) of diesel vehicle fleets. Together with diagnostic and failure prediction processes, it composes a data-driven vehicle maintenance structure that helps a fleet manager making decisions on a particular breakdown. The monitoring or controlof lubricating oils in diesel engines can be carried out in different ways and following different methodologies. However, the list of studies related to automatic lubricant analysis as methods for determining the degradation rate of automotive diesel engines is short. In this paper we present an intelligent data analysis from 5 different vehicles to evaluate whether the variables collected make it possible to determine the operating condition of lubricants. The results presented show that the selected variables have the potential to determine the operating condition, and that they are highly related with the lubricant condition.We also evaluate the inclusion of new variables engineered from raw data for a better determination of the operating condition. One of such variables is the kinematic viscosity which we show to have a relevant role in characterizing the lubricant condition. Moreover, 3 of the 4 variables that explaining 90$$\%$$% of the variance in the original data resulted from our feature engineering.