This is a review paper of studies that have employed the functional resonance analysis method (FRAM). FRAM is a relatively new systemic method for modeling and analyzing complex socio-technical systems. This review aims to address the following research questions: (a) Why is FRAM used? (b) To what domains has FRAM been applied? (c) What are the appropriate data collection approaches in practice? (d) What are the deficiencies of FRAM? A review of 52 FRAM-related studies published between 2010 and 2020 revealed that FRAM-based models can be used as a basis for improving safety management, accident/incident investigation, hazard identification/risk management, and complexity management in complex socio-technical systems. The outcomes also showed that healthcare was the most common domain that employed FRAM (31% of the investigated studies). The results of exploring data collection methods indicated a mixed method (interview, focus group, observation) was employed in 52% of the analyzed studies, and the accident investigation report was the most popular approach in aviation-related studies. An investigation of the deficiencies of the FRAM showed that it should be upgraded by exploiting supplementary methods to enhance its analytical and computational capacity to help risk analysts and safety managers in complex socio-technical systems. K E Y W O R D S accident investigation, complex socio-technical systems, complexity management, functional resonance analysis method (FRAM), hazard identification, safety management 1 | INTRODUCTION 1.1 | Background Complex socio-technical systems consist of some subsystems and subactivities linked in known or unknown ways (Hollnagel, 2012a, 2012b). Examples of socio-technical systems include healthcare, aviation, manufacturing, power industry, and automotive (Soliman & Saurin, 2017). They are inherently complex, nonlinear, uncertain, and dynamic (Jensen & Aven, 2018). Complex relationships between humans and their environments, including technologies and organizations, show that safety is not a linear and straightforward process in such systems (Grant et al., 2018). In the Safety-I perspective, the focus is on reducing adverse outcomes, such as accidents, incidents, and near misses (Hollnagel, 2018). The core idea of the established techniques for analyzing risks and accidents in the Safety-I approach is based on event chains: unexpected outcomes and potential accidents cannot be anticipated by considering event chains or possible