Abstract: Prognostics and health management (PHM) is crucial for ensuring the safe operation of machinery, improving the productivity and increasing economic benefits. High-quality life-cycle data, as the basic resource in the field of PHM, are able to carry the key information which reflects the complete degradation processes of machinery. However, due to the high costs in data acquisition and insufficient development in storage and transmission technology, typical life-cycle data is extremely scarce, which limits the theoretical research and engineering application of PHM for machinery. In order to solve this dilemma, accelerated life tests of rolling element bearings are carried out by Prof. Yaguo Lei's research group from School of Mechanical Engineering, Xi'an Jiaotong University (XJTU) and the Changxing Sumyoung Technology Co., Ltd. (SY), Zhejiang. These tests lasted for two years and the acquired datasets, i.e., XJTU-SY bearing datasets, have been publicly released for all PHM researchers. The XJTU-SY bearing datasets contain run-to-failure vibration signals of 15 rolling element bearings under three different operating conditions. These datasets have high sampling frequency, large amount of data, abundant failure types and detailed recording information. Accordingly, these datasets not only provide fresh "data blood" for PHM and promote the research of fault diagnosis and remaining useful life prediction, but also are able to help to improve intelligent maintenance decision making in industry.