While a number of valuable resources identify basic design options for building agent-based models (ABMs) (e.g., Gilbert, 2019; Rand and Rust, 2011), they do not offer clear guidance on the overarching concepts required for generating emerging outcomes. To encourage more researchers to engage in formally studying dynamics using ABM, we need to offer thorough design directions. We address this need by illuminating core design concepts. In this paper we offer four core conceptual distinctions in the design of ABMs that drive nonlinear relationships: (1) choosing a structure of relationships among the agents that create local impacts; (2) including opposing forces that constrain each other’s effects; (3) defining the sequencing of agents’ interactions and agents’ decision rules that alter dynamics; and (4) identifying boundaries to create asymmetry in agents’ interactions. Through a review of ABMs, we identified these concepts and developed criteria for making choices within them. We bring the concepts to life by illustrating them with examples from classic ABMs from several disciplines and from more recent management studies. Finally, we offer suggestions for moving from theories or phenomena to common design choices and provide key questions that researchers can use to guide the design of their ABMs to produce emerging outcomes.