Molecular epidemiology is moving from an era of single marker studies to omics-driven population-based studies, made possible by recent rapid technological advances in high-throughput laboratory assays. The use of these assays in large, observational, epidemiologic studies poses major methodological challenges. The traditional, reductionist approach that most epidemiologists have been taught is no longer sufficient with the introduction of omics data into epidemiology. However, basic epidemiological principles, such as confounding, must still be addressed. In order to successfully implement these tools into molecular epidemiology, scientists in the required disciplines must be able to effectively communicate. We propose an interdisciplinary, three-level, problem-based approach to training in molecular epidemiology. This approach requires interdisciplinary teaching teams that appeal to students coming from diverse backgrounds. First, concepts from each field necessary for molecular epidemiology must be addressed at a basic level, with a specific focus on biological, epidemiological, and statistical concepts. Second, state-of the art methods applied in molecular epidemiology should be covered (e.g., omics technologies, smart study designs, statistical modeling, and biological interpretation). Third, the application of omics technologies in population studies calls for awareness of issues related to biobanking, ethics, legislation, and public health. Problem-based learning is a useful tool for teaching when students need to apply complex concepts to complex situations as is the case in molecular epidemiology. Using this approach, students from diverse backgrounds work together to analyze problems, discuss, and exchange knowledge in modern molecular epidemiology. Training tools characterized by this approach will shape the molecular epidemiologists of the future. Environ. Mol. Mutagen. 54:611-615, 2013. © 2013 Wiley Periodicals, Inc.