The viscosity of lubricating oil is the core characteristic of its mechanical properties,which is influenced by temperature directly.To determine the relationship between viscosity and temperature,the viscosities of six types of lubricating oils were measured by experiments,and the trend between viscosity and temperature was analyzed.Several viscositytemperature correlation models were applied to fit the measured data.It is found that the trend between viscosity and temperature is similar to exponential function in general.The single logarithmic method can’t eliminate the exponential trend of viscositytemperature correlation completely,so the fitting residual error is larger.The MacCoullWright model which uses the double logarithmic method has a high fitting precision and it has a wide field of application.The triple logarithmic method can’t improve the fitting precision in contrast with the double logarithmic method,besides,the calculation process is complicated.In the case of low viscosity of lubricating oil such as high temperature condition,the data exceed the logarithmic function definitional domain,hence it restricts the application range of this method.