Context and setting Medical educators are continually challenged to add new topics to the curriculum. Calls for teaching about medical error and the science of improving patient safety come from multiple accrediting and credentialing bodies, as well as patient advocacy groups. Medical educators are faced with the challenge of how and where to begin teaching the scope of the issue, the necessary vocabulary and the ‘basic science’ of system improvement amidst an overcrowded curriculum. The second year of basic science at the College of Human Medicine, Michigan State University is delivered using a problem-based learning (PBL) curriculum. Organised in 10 organ or system domains, patient cases integrate basic science content and are supplemented by references, stimulus questions, and some lectures and laboratory sessions. Why the idea was necessary There is consensus that the teaching and assessment of medical error and patient safety knowledge, skills and attitudes should begin as early as possible so that learners are ready to apply these principles to their clinical experiences. Licensing organisations are adding questions on these topics to their examinations. The patient cases in our PBL curriculum have already provided the context for ethics, evidence-based medicine and other topics, the content of which is relevant to multiple PBL domains. Using these cases to provide context for medical error and patient safety-related content seemed logical. What was done Since 2008, the Cardiovascular Domain Curriculum Development Group (CDG) has added foundational material on medical error and patient safety science to several patient cases. An introductory lecture in the first week of the domain covered the scope of the issue, basic principles of error genesis, failure mode effect analysis, human factor engineering and root cause analysis (RCA). A pre-test of 10 multiple-choice questions preceded the lecture. Hospital patient safety goals (relating to communication error, medication reconciliation, health care-associated infection) guided the interpolation of medical errors into the existing cases. One ‘patient’ died as a result of a hand-off error. Stimulus questions for small-group discussion were added to cases with error and references were provided for independent study. On the last day of the domain, small groups watched a mock RCA. Five of the pre-test questions were placed on the final domain examination. Evaluation of results and impact Overall performance on the domain examination did not differ from that in previous years. For the five questions that appeared in both the pre-test and the domain examination, the mean student score was 50% in the pre-test and 79% in the domain examination. Mean student rating of the lecture was 3.6/5.0. Subsequently, a Year 3 student commented: ‘At the time I was not sure why we were learning this. We did an RCA on one of my patients this [clerkship] year … I knew what they were talking about and why it was done.’ One faculty member commented: ‘Thanks for adding this important material.’ The addition of foundational medical error and patient safety material to a PBL domain placed this material into a patient context, in which it was well received. Students demonstrated improvement in their knowledge base. Results were presented to our curriculum committee and a decision was made to add medical error and patient safety material to cases in every one of the PBL domains.