Rolling bearing reliability is often affected by the processing quality, however the life reliability or operating reliability do not involve the study of rolling bearing processing quality assessment. In order to avoid bearing failure caused by bearing processing quality, it is necessary to evaluate the reliability of the bearing quality assurance capability. The difficulty in reliable evaluation of bearing's quality assurance capability is how to quantify the uncertain relationships between the bearing quality and its influence factors, and then determine the weight of different influencing factors. Therefore, in this paper, a novel method for determining the weight of bearing quality-influencing factors based on bootstrap maximum entropy method and similarity method is proposed, and then the quality-achieving reliability model is established. The experimental results show that the proposed method can effectively quantify the relationship between the bearing quality and its influencing factors, and accurately assess the bearing quality assurance capability and bearing quality processing level under the condition of a small number of experimental bearings. Compared with other quality-achieving reliability method, the proposed method is more effective.