Catalyst deterioration caused by phosphorus-containing engine oil additives was investigated using a variety of engine oil blends in a steady-state engine-dynamometer test. The reductions in hydrocarbon and carbon monoxide conversion in the 200-hour test were related to two parameters: 1) the quantity of phosphorus in the oil added to the engine, and 2) the amount of phosphorus on the catalyst at the end of the test.Catalyst degradation relative to the first parameter differed from that relative to the second because the two parameters were not directly related. Specifically, catalyst conversion efficiency decreased nonlinearly with the amount of oil-derived phosphorus added to the engine, but linearly with the amount of oil-derived phosphorus found on the catalyst. A higher percentage of phosphorus added to the engine was found on the catalyst with oils containing tricresylphosphate (TCP) than with oils containing zinc dialkyldithiophosphate (ZDP). For a fixed amount of phosphorus on the catalyst, however, oils containing ZDP were more harmful to catalyst performance than oils containing TCP.For most oils containing ZDP, phosphorus was observed to accumulate in sump oil during engine operation. As a result, less phosphorus than that added to the engine was responsible for the decrease in catalyst conversion efficiencies observed with these oils.