Rear axle tapered roller bearing wear has been studied in bench and car tests, using both commercial and experimental gear lubricants. In bench tests, magnitude of wear was affected by both lubricant additive treatment and lubricant viscosity; increased wear resulted from reduction in viscosity with lead soap-active sulfur and Pb-S-Cl lubricants but not with P-S and Zn-P-S-Cl lubricants. Substantial decreases in wear from one bearing lot to subsequent lots apparently resulted from mechanical improvement in the bearing manufacturing technique. These wear decreases, for a given bearing speed and load combination, were accompanied by decreased temperature and thickness of roller-end surface film formed from the lubricants. In rear axle car tests run as far as 94,000 miles, P-S additive package blends having viscosities as low as 4 cs at 210 F performed satisfactorily, although bearing wear was somewhat greater than that with P-S blends having viscosities of 9 cs at 210 F.