Diagnostic errors cause substantial harm to patients but remain largely unappreciated within the patient safety movement in health care.1 The National Academy of Medicine defines a diagnostic error as “failure to (a) establish an accurate and timely explanation of the patient’s health problem(s) or (b) communicate that explanation to the patient,” but this universal definition can be difficult to apply in a fast-paced critical care environment, such as the intensive care unit.1 An operational definition adopted by some researchers is that a diagnostic error is present if there is unequivocal evidence of a missed opportunity during the diagnostic process to establish a timely and accurate diagnosis.2 This is especially applicable in settings such as the neonatal intensive care unit, where the length of stay is often prolonged and most diagnoses will be discovered before discharge. The key to determining if a diagnostic error occurred is to establish whether there was an opportunity to make the diagnosis sooner. However, when performing retrospective evaluations for the presence of diagnostic errors, researchers operate without a direct association with either the treating clinician or patient. This leads to the following quandary: do researchers have an ethical obligation to inform patients or clinicians when they discover previously unrecognized diagnostic errors? In this Viewpoint, we review the current literature on the ethics of disclosing adverse events, such as diagnostic errors, and then formulate a framework to guide the researcher in this scenario.