The primary role of COPE (the Committee on Publication Ethics) is to provide advice and guidance to journal editors and publishers on ethical issues. COPE also expects its member journals to follow its Code of Conduct (1). The Code emphasizes that editors have a duty to pursue cases of suspected misconduct but makes it clear that such cases should be investigated by the researchers’ institutions, not by the journal.COPE therefore recognises the responsibility of research institutions (eg. universities and hospitals) to investigate cases of suspected research or publication misconduct such as data fabrication or plagiarism. COPE also emphasizes that journal editors should take responsibility for everything published in their journals, and recognises that this often requires liaison with institutions, especially in cases of suspected or alleged misconduct.Although journals should not generally attempt to investigate cases of suspected misconduct, editors are often the first people to become aware of possible problems. Suspicion of misconduct, or of honest errors, may surface during peer review or be raised by readers. When this happens, editors often need to refer the matter to the authors’ institution(s) and request an investigation or clarification. Similarly, if an institutional investigation reveals problems in published work, editors need to respond appropriately, for example by issuing a retraction or correction.Even when serious misconduct is not suspected, editors may seek clarification or arbitration from institutions, for example in cases of disputed authorship or if questions arise about the ethical review or conduct of research. Therefore, it is important that journals and institutions have effective channels of communication and the ability to cooperate.However, despite the importance of journals and institutions working together, COPE is aware that such cooperation does not always take place. COPE member journals can bring challenging cases to its quarterly Forum meetings to seek advice from other members. The anonymised cases are entered into a database which can be searched from the COPE website. Such cases cannot give an indication of the true frequency of problems, because editors probably bring only their most difficult cases to COPE. Nevertheless, an analysis of cases in which editors had tried to work with institutions showed a worrying number of problems, ranging from unclear communications to institutions not responding at all (2). Other sources also suggested that journals do not always respond appropriately when contacted by institutions (3).In the light of these findings, COPE therefore decided to develop guidelines on cooperation between journals and research institutions on research integrity cases (4). This document is based on COPE’s experience of the problems editors may face, and on wide consultation with institutions throughout the world and those responsible for research integrity and investigating cases of suspected misconduct.The key assumption underlying the guidelines is that, while journals should be responsible for the reliability of everything they publish, institutions should be responsible for the conduct of their researchers. In cases of proven serious misconduct, such as data fabrication or major plagiarism, journal and institutional policies are usually well aligned. For example, in such cases, the institution would normally require researchers to retract published articles and editors would also wish to do this to prevent readers from being misled by unreliable publications. However, the guidelines make an important point that editors’ responsibility to safeguard readers means that they may use definitions of misconduct (or inappropriate publication activities) that are stricter than those used by institutions. Problems can therefore arise if a journal asks an institution to investigate behaviours such as duplicate publication, which clearly breach the journal’s policies and may mislead readers (and break copyright rules), yet are not considered to be serious misconduct by an institution. In such cases, some junior researchers might be found not guilty of misconduct by their institution, but the journal may seek to retract the second publication. Because of the different responsibilities of journals and institutions, and recognising that they may, quite legitimately, use different definitions of misconduct, the guidelines do not attempt to define misconduct but focus, instead, on the need to correct or retract publications that are unreliable, whatever the cause (including honest error).The guidelines focus on how journals and institutions should respond to cases of suspected misconduct, because we felt little or no guidance was available on this topic. However, although not addressed in the current guidelines, COPE also recognises the importance of education, training, and other processes aimed at preventing misconduct or detecting problems at an early stage.The guidelines stress the need for effective communication between journals and institutions but also recognise the need for confidentiality in certain circumstances or at certain stages of an investigation. One recommendation we particularly hope that institutions will follow is to identify a senior staff member (or team) with responsibility for research integrity. COPE believes this will greatly facilitate communication between journals and institutions and also encourage researchers to seek guidance about their own work and to come forward with suspicions of misconduct or poor practices within the institution.COPE hopes that journals will promote and publicise the new guidelines, and appreciates the efforts of the Saudi Journal of Anaesthesia in doing this. We also hope that researchers will discuss the guidelines in their institutions and work to incorporate them into institutional policies on research integrity. We also welcome comments and will review and revise the document as required. We hope that these guidelines represent a first step in closer collaboration and cooperation between journals and research institutions which may lead to better understanding and, ultimately, to higher levels of research and publication integrity and swifter, more efficient resolution of misconduct cases.