Most analyses of plagiarism focus on published content and do not report on the prevalence of plagiarism in submitted articles. Fears over large‐scale plagiarism, particularly in articles submitted by authors for whom English is a second language, have only been investigated in small publishing communities or using duplication‐checking analysis, which does not separate legitimate from unacceptable duplication. This research surveyed journal editors from around the world to ascertain recent (past year) experiences of plagiarized and/or duplicated submissions. We then compared their experiences to their assumptions about global levels of plagiarism. The survey received 372 responses, including 119 from Asian editors, 112 from European editors, and 57 from editors in North America. The respondents estimated that c.15% of all submissions contained plagiarized or duplicated content, although their own experiences were in the range of 2–5% of submissions. Of the respondents, 42% reported no incidence of plagiarized or duplicated submissions in the past year. Asian editors experienced the highest levels of plagiarized/duplicated content, although most of the problem articles were resolved, indicating that most of the identified duplication constituted relatively minor problems, rather than fraudulent plagiarism.