Healthy and diseased tissues—including cancer—are adaptive ecosystems that are driven by individual cell behaviors, interactions, and biophysical constraints. Many key cell behaviors and interactions—particularly tumor-immune interactions—are fundamentally discrete, requiring discrete modeling approaches. In this talk, we introduce agent-based modeling (ABM) in cancer biology, highlight major on-lattice and off-lattice approaches, and explore examples in cancer hypoxia and cancer immunology using PhysiCell: an open source ABM framework (and software ecosystem) developed by my group. We close with next-generation ABM approaches to accelerate model development, improve multidisciplinary communication, increase access to mathematical modeling outside the traditional math/physics/CS community, and improve reproducibility and reusability of agent-based models.