The aim of this study was to analyze the features of retracted articles by authors affiliated with hospitals in mainland China. We searched the PubMed, Web of Science, and Retraction Watch databases for retractions and identified the following characteristics of each retracted article: publisher, open access status, impact factor of the journal that retracted the article, any PubPeer comments recorded before the retraction, status of the hospital where the authors worked, and any response to the retraction from the authors. We found 521 retractions, primarily by authors at grade A, third-level hospitals located in a limited number of regions of mainland China, and found that the journals that had published and later retracted the articles tended to have a medium to high impact factor. The main reasons for retraction were data manipulation, fabrication, or fraud; errors made by the authors; or plagiarism. Few of the retracted publications had PubPeer comments before their retraction. This is the first report to focus on retracted research coming out of hospitals in mainland China. The large number of retractions for Chinese hospitals is worrying. The results suggest that some retractions were related to third parties that provided editorial and other services.