Abstract: The service life of products is significant for ensuring stable and reliable performance of these products, such as spacecraft, military equipment and electronic equipment. Such service life is generally evaluated by measuring the working duration of products before losing efficacy or breaking down through tests under the same working condition with the real working condition. However, with the rapid increasing improvements in reliability and of products, the service life of products had been lengthen remarkably, the corresponding time and financial costs of the conventional simulated products life tests were continuously increasing, and such costs were becoming too high to be accepted. On this basis, the concept of accelerated test was suggested as a potential solution in which appropriate equivalent of products service life should be selected. The tests were accelerated by adjusting the changing rate of the equivalent, so that the test durations were shortened under a certain equivalent. In 1980s, the Accelerate Life Test (ALT) and the Accelerate Degradation Test (ADT) based on ALT, which estimated the reliability and service life of products under common stress levels through the accelerate test data under higher stress levels were suggested. ALT and ADT were the most common accelerated test methods, in which the failure possibility function were regarded as the equivalent. In other accelerated tests, the selected equivalent also included surface topography and condition parameters. As for accelerated wear test, there were numerous models for quantitative prediction of wear, which was a typical form of the degradation process of materials, the equivalent of wear could be selected based on these quantitative wear models. This paper aimed to propose an accelerated wear test method based on an accurate quantitative wear model. The investigations were conducted based on friction and wear with experiment condition of surface contact and mixed lubrication by the standard apparatus for typical mechanical seal material pair 9Cr18Mo/M234A0. The structure of tested seal ring was modified to reduce temperature rise and strengthen the fluid dynamic pressure effect. In this paper, the accelerated wear model was established with the entropy increase as equivalent of wear, based on the dissipation wear model which considered the wear as an irreversible thermodynamic process. The dissipation wear model was verified with the mentioned seal pair, revealing a linear relationship between wear rate and entropy increase. By modifying the entropy increase rate under the same entropy increase, the wear test duration was shortened. On this basis, this accelerate wear model was verified with equivalent wear tests with long time/low entropy increase and short time/high entropy increase. This model was also compared with accelerate wear model based on Archard model. The result of which revealed that wear test of 200 min duration could be equally replaced by a wear test of 10 min duration with an error of 2.22%, and the maximum relative error of other time was 4.75%. The feasibility and the accuracy of this accelerate wear model was proved preliminarily by the test results of this study. Further investigation revealed that the wear mechanism was invariable during the tests in this paper.