Background: Perhaps the most common of all software engineering activities is the modernization of software. Unfortunately, during such modernization often leaves behind artifacts that are difficult to understand for those other than its author. Thus, the Object Management Group (OMG) has defined standards in the modernization process, by creating the concept of Architecture-Driven Modernization (ADM). Nevertheless, to the best of our knowledge, there is no a systematic mapping study providing an overview of how researchers have been employing ADM. Thus, we assert that there is a need for a more systematic investigation of the topics encompassed by this research area. Objective: To describe a systematic mapping study on ADM, highlighting the main research thrusts in this field. Method: We undertook a systematic mapping study, emphasizing the most important electronic databases. Results: We identified 30 primary studies, which were classified by their contribution type, focus area, and research type. Conclusion: This systematic mapping can be seen as a valuable initial foray into ADM for those interested in doing research in this field. More specifically, our paper provides an overview of the current state of the art and future trends in software modernization area, which may serve as a road-map for researchers interested in coming up with new tools and processes to support the modernization of legacy systems.