Managing non-communicable diseases requires policy makers to adopt a whole systems perspective that adequately represents the complex causal architecture of human behaviour. Agent-based modelling is a computational method to understand the behaviour of complex systems by simulating the actions of entities within the system, including the way these individuals inﬂuence and are inﬂuenced by their physical and social environment. The potential beneﬁts of this method have led to several calls for greater use in public health research. We discuss three challenges facing potential modellers: model speciﬁcation, obtaining required data, and developing good practices. We also present steps to assist researchers to meet these challenges and implement their agent-based model.