Laboratory engine tests have beets used to evaluate the performance properties of engine oils for more than four decades and have been progressively modified to meet the lubrication challenges from new engine designs, operating conditions, fuels and emission controls. Engine sequence or bench laboratory tests are an important first step in the development of passenger car crankcase oils. These tests are correlated with reference oils of conventional additive chemistry with known field performance. When using conventional additive chemistry, sequence tests usually provide an excellent correlation with field performance. However, when new or novel chemistry is employed, laboratory testing may not be an adequate indicator of field performance. In these situations, actual field tests must be conducted. There are also other reversals or unexpected results which further require field tests as the final proof of performance. This paper presents specific examples of field test performance which were not predicted by laboratory engine testing. Additionally, field test results addressing new gasoline engine performance requirement's (API SG) will be discussed.