Compared to real-world exercises and drills, computational simulation provides an inexpensive, safe, and time-efficient means of carrying out what-if scenarios to examine alternative strategies for a given situation. Furthermore, it enables investigations and predictions into how the complex dynamics of a real-world system are likely to be affected by changes to internal and/or external factors. Over the past two decades, the development and application of computational simulation models have rapidly increased in a wide range of domains. Crisis and Disaster Management (CDM) is one domain in which this has been witnessed, driven by the significant increase in the number of large-scale and unprecedented emergency situations and the need to better manage them and lessen their impact. These include areas such as pandemics, forest fires, floods, and wide-scale population movements (e.g., refugees). Computational simulation of natural and man-made crises and disasters offers opportunities to bring together stakeholders, including academics and practitioners from a variety of backgrounds, in the design and implementation of pragmatic and fine-grained models applicable to the various stages of the emergency management life cycle (preparation, response, mitigation, and recovery). Such collaborative efforts have delivered great benefits via sophisticated computational simulation models. Indeed, these models have improved trustworthiness, usefulness, usability, and validity. Their application has seen growing acceptance and relevance, thus leveraging their results to make a practical difference. Consequently, the simulation results produced can contribute to providing a sound basis for “better” decisions regarding how to manage crises and disasters. However, there are still many challenges to be faced: • incorporating realistic human behavior into CDM simulation models, • increasing acceptability of simulation models by citizens (democratization) and other stakeholders, • developing a hybrid approach by coupling simulation modeling (e.g., agent-based, virtual reality, systems dynamics, participatory approaches) with non-simulation modeling (e.g., optimization, social media analytics). This helps to apply the separation of concerns principle and thus better deal with the complexity of the universe to be modeled, • creating multi-scale models by combining micro-level (e.g., citizen behavior) and macro-level (e.g., crowd behavior, norms) representations. The challenge is to understand how they influence each other and how macro-level phenomena emerge from micro-level behaviors in the CDM context, • explaining and predicting phenomena from real CDM data, • validating CDM simulation models and dealing with quality issues, such as robustness, scalability, efficiency, or reliability. This special issue focuses on current, real-world problems (pandemics, humanitarian crises, natural and man-made disasters), highlights the use of computer simulation in crises, and addresses the challenges presented earlier. Notably, many of the papers in this issue use an agent-based approach, which shows not only its popularity but also its usefulness. This approach has numerous advantages: It is possible to model intelligent and autonomous individual behaviors; it can incorporate high-level interaction protocols (e.g. negotiation, contracting, and so on); it can easily take into account organizational abstractions, such as roles and groups; and it can capture emergent phenomena resulting from the micro-level interactions between agents.