Abstract Electronic medical records (EMRs) allow for the creation of "test" patients that can be used for training new users, testing changes before they go "live", simulating the placement of orders, or for a range of similar applications. Test patients are commonly utilized in EMR validation, support, or “playground” environments that might use a copy of the production EMR database but do not connect to the official medical record to prevent billing errors or patient harm. However, there are a variety of cases where it is useful to have test or anonymous patients within the EMR production environment as well. One example of this is infectious disease testing for employee blood-borne pathogen exposure (e.g., needlestick injury) that may need to be stored in a de-identified or restricted manner to comply with workplace privacy regulations. There are also instances, generally in emergency situations, where a medical record is created for a patient about whom little identifying information is known. In this example, these so-called "anonymous" patients are assigned random or arbitrarily pre-determined birthdates (e.g., 1/1/1900) and on occasion have no sex assigned in their chart. This poses a variety of challenges from a laboratory standpoint as it can be difficult to provide accurate normal ranges for laboratory results that are age and/or gender specific. Additionally, this has on occasion resulted in an inability to identify appropriate blood products as these recommendations are similarly based on age and gender. Another layer of complexity comes as additional information regarding the patient is made available, and health care information services merges data from a previously anonymous account to a specific patient record. Such mergers can result in cancellation of pending orders, errors in reporting laboratory results, or other complications depending on the specific process utilized. Surprisingly, there is little to no literature regarding the use cases for these “test” and “anonymous” patients or the challenges associated with their existence in the EMR. To explore this topic, we identified three classes of patients in our institutional EMR: true virtual “test” patients, “anonymous” patients that were subsequently identified, and “anonymous” patients that were never identified. Basic characteristics of these patients (sex and date of birth) and available laboratory results were compiled. Along with this, we present a series of instructional cases adapted from actual patient safety events at our institution involving anonymous patient records. These illustrative cases highlight the utility of these “test” and “anonymous” patients, as well as the challenges these pose on an institutional and individual level. Lastly, we provide recommendations on how best to manage similar scenarios that may arise.