Two programs were conducted to study the relationships between engine oil rheology and crankshaft bearing wear. A Chassis Dynamometer test of four oils in four cars was used to explore and define the key variables affecting bearing wear. These results were used to design a Field Test of nine oils in 45 taxicabs in New York City. The test oils (SAE OW-20 to 20W-20) were formulated to measure the effects of viscosity, viscosity index improver, and detergent inhibitor package. Bearing wear tended to be either low and unremarkable or very high, particularly in the thrust bearings. Oil performance was best expressed as the frequency of excessive wear, rather than by quantitative wear measurement. There were many instances of very high wear in cabs operated with the lowest viscosity oils but none in cabs with higher viscosity oils. Non-Newtonian oils appeared to provide slightly more protection than Newtonian oils of the same HTHS viscosity, and a higher quality adpack also appeared to provide benefits. However, these factors were secondary to the viscosity of the oil. HTHS viscosity was a better predictor of bearing wear performance than oil film thickness.