The irrationality changes of motor oil has caused various problems, such as excessive consumption of lubricating oil, high maintenance costs and enduring environment pollution. Multiple road tests of 7 civilian cars were studied in this paper, which were operated under normal urban conditions. The oil samples during experimental process were regularly collected and the oxidation onset temperature (OOT) was tested by using PDSC method. The multiple linear regression method was used to establish the relationship model between the OOTs of the lubricants and vehicle parameters, including service mileage and service time based on the 312 samples. The results showed that the replaced lubricant exhibited high OOTs and still retained excellent oxidation stability, which suggested the current oil drain interval is conservative. Therefore, the level of oil drain interval in our country need to be further improved. The data analysis showed that the established models were concise with good fitting and prediction accuracy. Meanwhile, the models can be used for end users to judge the changing of lubricating oil easily and conveniently.