Health economic (HE) studies comprise a variety of analytic approaches and utilize disparate data sources. Study objectives can range from determining the economic burden of a particular disease to evaluating the budgetary impact of a novel therapy. Data can be derived from payer databases, government sources, peer-reviewed literature, and myriad other sources. Health economic studies are conducted by different types of researchers working at various types of institutions. One of the more compelling types of HE study designs is a cost-effectiveness or cost-utility analysis “piggybacked” on a randomized controlled trial of an investigational pharmaceutical, biologic, or medical device or on an observational study of the real world use of a marketed product. While piggyback studies can be conducted across multiple institutions, they are perhaps most efficiently conducted in totoby a singular research organization. Building on lessons learned from actual studies, this presentation will elucidate best practices for design and execution of HE analyses piggybacked on Phase 2/3 trials or observational studies. Best practices presented will range from study initiation through analysis and dissemination of findings. Beyond electronic data capture (EDC) systems historically used in earlier phase clinical trials, piggyback studies may utilize electronic health record (EHR) data, claims data, patient reported outcomes (PROs), and other data. These data are companied by new sets of analytic, logistic, and regulatory challenges that must be addressed through all stages of study design and execution. Negotiation of study site contracts and payments and the development of informed consent procedures must reflect regulatory expectations and requirements for both clinical trials and HE studies. Database merging and validation, including the juxtaposition of coding for registrational studies with coding for medical billing, require particular consideration. Finally, simultaneous planning for, and execution of, analysis and reporting of results are essential.