An engine condition monitoring research method was proposed based on orderly sample clustering and Fuzzy theory.Orderly sample clustering method was applied to divide oil ferrography data into several categories and the engine condition monitoring was realized.Fuzzy theory method was applied to find out the location of the abnormal friction pairs of engine in combination with atom emission spectrum analysis data.The method was adopted for the conditioning monitoring of the diesel engine Cummings 6BT59,the possible abnormal wear position was determined,and the results were in accordance with the actual inspection results of engine dissemble.It is proved that the above research method is suitable for identifying the abnormal wear portion of engine.