This paper presents the findings from a proof-of-concept study that was conducted to examine whether engines and vehicles equipped with onboard diagnostic systems could provide data for optimizing fleet preventive maintenance practices. The study investigated the development of a statistical approach for recommending oil changes in the Texas Department of Transportation's fleet, based on engine data collection and oil sampling analysis for a sample of heavy-duty dump trucks. The study also investigated whether predictive intervals could improve preventive maintenance practices and save money. The engine parameters that were logged included engine speed, oil temperature, engine load, coolant temperature, and engine oil pressure; and the oil parameters tested included viscosity, oxidation, nitration, total acid number, total base number, wear metals, soot, and fuel dilution. The findings indicated that there were very low levels of oil degradation, attributable to the engine operations, which were observed to be predominantly low-load operations with a lot of idling. The findings support potential changes to fleet maintenance and management practices from an economic and environmental perspective.