Since many countries have lengthy periods between initial marketing authorisation and country approvals and reimbursement, the number of early access programs (EAPs) (also known as compassionate use or named patient programmes) being initiated by pharmaceutical companies are increasing to bridge the gap between clinical trials and market-uptake. Through EAPs, patients who have either already benefitted from investigational agents or who demonstrate unmet need can receive promising new treatments. In addition to providing early treatment access, EAPs offer a unique opportunity to evaluate clinical and safety outcomes outside the trial setting, without the constraints of strict inclusion and exclusion criteria. Data collected in EAPs are viewed by some as proxies for real-world use since the potential benefits of an investigational treatment can be observed in a wider range of populations. Available EAP guidelines regarding acceptable data collection beyond safety outcomes is limited. There is a mix in published EAPs with some collecting data in the program with others collecting data via chart review after programs close. While the United Kingdom’s Early Access to Medicines Scheme (EAMS) and France’s Temporary Authorisation for Use programs appear to allow data collection beyond safety, the directives remain vague. Generally, the relevance of EAP data collection for marketing authorization appears to be limited, except for EAMS guidance, which indicates that data generated in EAMS can be used to facilitate NICE Technology Appraisals. Data collection approaches must be scientifically robust, practical, and ethical. Careful consideration for incorporating data elements into EAPs, how data will be used and challenges this may impose (that could negatively impact the program) is essential. Once the data elements are delineated, stakeholder feedback and approval from the regulatory authority that governs the EAP are needed. Secondary feedback and endorsement from clinician and hospital governing bodies is also recommended.