PURPOSE: The purpose of this workshop is to highlight best practices to improve observational research performance through 1) a methodological overview of algorithm development and validation in electronic health records (EHR) and healthcare claims-based databases, 2) an understanding of novel approaches using text-mining and natural language processing (NLP), and 3) case studies of algorithms developed to identify multiple sclerosis (MS) and its clinical subgroups to generate real world evidence (RWE). By offering an interactive forum, participants will be able to identify and share their successes and challenges in algorithm development and validation. DESCRIPTION: We will first present an overview of algorithm development methodologies in EHR and healthcare claims databases. Based on the nature of the research question and study outcomes of interest, we will also address the framework for identifying databases and coding systems with a focus on feasibility, data capture, and data quality. The algorithms’ development process, evaluation measures, and best optimization practices will be shared through examination of case studies that identify MS patient cohorts, subtypes, relapse, and disease activity status. The second topic will focus on novel text mining and natural language processing approaches to enhance performance. NLP-based and traditional medical chart review validation techniques will also be compared. Finally, we will provide insight into the potential benefits and obstacles of using algorithms to generate RWE for the value and safety of medicines. The audience will be engaged in an interactive discussion regarding their successes and challenges in using validated algorithms with contemporary data sources. This workshop will be valuable for healthcare researchers wanting to gain greater awareness and insight on how algorithms can be used to support unmet need and effectiveness studies in health economics and outcomes research (HEOR).