The growing array of record-keeping laws, ethical standards, and professional guidelines has created controversy and confusion. Clinicians struggle with what to leave in, what to leave out, how to handle records securely, when to respond to requests for records versus when to refuse, and so on. This article focuses on 5 challenging areas: confidentiality; informed consent; the state, the law, and legal requirements; third-parties; and the implications of research findings for record keeping. It discusses published claims, critiques, proposals for change, and research reports, particularly those of Bemister and Dobson (2011, 2012); Castonguay (2013); Christie, Bemister, and Dobson (2014); Furlong (2013); and Mills (2012). It emphasizes the potential problems with any 1 size fits all approach and the difficulties in creating sensible regulations that do justice to the diversity of values, contexts, cultures, and theoretical orientations.Keywords: record keeping, ethics, confidentiality, electronic health records, clinical recordsResumeLa variete croissante des lois, des normes deontologiques et des lignes directrices professionnelles a suscite controverse et confusion. Les cliniciens cherchent tant bien que mal a definir ce qui doit etre garde ou pas, a trouver la facon de tenir des dossiers de facon securitaire, a savoir quand acquiescer aux demandes concernant des dossiers ou quand refuser, ainsi de suite. Cet article porte sur 5 domaines difficiles a gerer : la confidentialite, le consentement eclaire, les exigences de l'Etat, de la loi et les exigences legales, les tierces parties et les implications des resultats de recherche en ce qui a trait a la conservation des dossiers. Sont discutes les allegations publiees, les critiques, les change- ments proposes et les rapport de recherches, en particulier les articles de Bemister et Dobson (2011, 2012), de Castonguay (2013), de Christie, Bemister et Dobson (2014), de Furlong (2013) et de Mills (2012). L'article met aussi en relief les eventuels problemes des solutions dites universelles et la difficulte d'etablir une reglementation sensee qui tient compte de la diversite des valeurs, des contextes, des culture et des approches theoriques.Mots-cles : tenue de dossiers, ethique, confidentialite, dossier de sante electronique, dossiers cliniques.Clinical records hold life-changing power. A record's facts, inferences, conclusions, gaps, inaccuracies, wording, and tone can affect whether a person keeps custody of a child, gets a security clearance, receives life-saving help in a crisis, or secures needed accommodations at work for a disability. When a record's security is breached, the content- diagnosis, medications, clinical history, and a patient's most sensitive and private information-may find its way to an array of people and organizations, perhaps exposing the patient to gossip, ridicule, identity theft, exposure of private information on social media, and worse. For example, after learning that an employee hates her job and boss, a company may reorganize so that the employee's position is no longer needed. If records of a clinical psychologist's own therapy leak out, it might damage the psychologist's alliance with his or her own therapist and could influence current and future patients' decisions to consult another therapist. A battered woman's husband may discover that she was seeing a therapist, despite his threat that he would kill everyone in their family if she did so.The field owes a debt of gratitude to Drs. Taryn Bemister, Suzanne Castonguay, Chelsea Christie, Keith Dobson, Allannah Furlong, and Jon Mills for bringing their expertise to bear on the diverse, sometimes conflicting demands, values, contexts, theoretical orientations, and other factors that make clinical records such a difficult area (Bemister & Dobson, 2011, 2012; Castonguay, 2013; Christie et al., 2014; Furlong, 2013; Mills, 2012). Their thoughtful presentations and exchange of views advance our understanding of the complex challenge record keeping poses and why this puzzle defies simple, easy solutions. …