Abstract: Objective  To systematically analyze the characteristics of articles retracted due to “paper mills” and explore effective governance strategies to uphold academic integrity. Methods  Based on data from the Retraction Watch Database (2010-2024), a total of 7 513 articles retracted due to “paper mills” were included. Descriptive statistics and categorical analysis were employed to examine disciplinary distribution, national characteristics, article types, retraction timelines, publisher sources, and retraction reasons. Results  Retracted articles were primarily concentrated in Business & Technology (48.3%) and Basic Life Sciences (38.1%). China (including Taiwan Province) contributed to 88.5% of global retractions. Research articles constituted the majority of retracted types (88.3%). The average retraction period was 761.5 d. Publishers such as Hindawi and IOP Publishing were major sources of retractions. Key contributing factors included third-party investigations, data irregularities, and peer review fraud. Conclusion  Retractions linked to “paper mills” exhibit features such as disciplinary concentration, geographical imbalance, publisher accountability deficits, and retraction delays. Multifaceted governance strategies—including reforming academic evaluation systems, strengthening journal review mechanisms, promoting international cooperation, and enhancing research ethics education—are essential to combat paper mill activities and safeguard the integrity of scientific research.

Key words: paper mills, retracted paper, academic misconduct, research integrity, academic evaluation, journal review