Bonding courses, also called leveling courses, are a universal feature of Roman architecture. This recognizable technique of the Roman builders spans the ancient empire from Great Britain to Turkey. Even though this technique is essentially ubiquitous, the full extent of its structural functionality remains in question. The literature is far from being unanimous. Some sources claim that bonding courses were originally used to give a level finish at the end of each stage of construction, resulting in potential lines of cleavage, and thus being detrimental in the design. Contradicting these ideas, others have suggested that bonding courses are advantageous because they help resist crack propagation and foundation settlement. The aim of this paper is to study the bonding courses from an engineering perspective and contribute to a better understanding of their structural functionality. This interdisciplinary work uses a combination of Finite Element Modeling (FEM) and Distinct Element Modeling (DEM) to simulate structural response of bonding courses to a variety of scenarios, including dead load, lateral loads, and differential settlement. Important properties of FDEM are a possibility to model discontinuous (jointed) structures and to implement an explicit time-step in the analysis. Enabling jointed structures to be properly simulated, this approach is proper for masonry where the stones are mechanically much stronger than the mortar joints. To ensure historical accuracy and rigorous engineering analysis, this project is realized in collaboration between the Civil and Environmental Engineering Department at Princeton University and the Institute for Mediterranean Archaeology.