The fate of rejected manuscripts that were originally submitted to the American Journal of Roentgenology (AJR) during the first 5 months of 1986 was investigated to learn whether, when, and where they had been published. AJR, a peer-reviewed journal of diagnostic radiology with a circulation of over 21,000, annually publishes about 500 papers and receives over 11,500 citations. MEDLINE searches conducted 45 to 54 months after the dates of rejection by AJR located 162 (64%) published papers out of a consecutive series of 254 manuscripts rejected by AJR, including 69% of the rejected major papers and 62% of the rejected case reports. The papers had been published in 30 different radiologic and 27 different nonradiologic journals. Most of these journals published fewer papers, had smaller circulations, and had lower impact factors (a ratio of citations received to papers published) than AJR does. The mean time lapse between rejection by AJR and publication in other journals was 15 months. The delay in publication was greater for papers published in nonradiologic and foreign journals than for papers published in radiologic and American journals. The results of this study indicate that rejection of a manuscript by a peer-reviewed journal such as AJR delays but by no means precludes publication. At least 82% of the major papers and 70% of the case reports that are submitted to AJR are eventually published, either in AJR or elsewhere. Because a scientific paper represents not only many hours of writing and manuscript preparation but also a great investment of research time and resources, authors are reluctant to abandon rejected manuscripts. In the majority of cases, submission to other journals gains acceptance and publication.