Ecologists and decision makers need ways to understand systems, test ideas, and make predictions and explanations about systems. However, uncertainty about causes and effects of processes and parameter values is pervasive in models of ecological systems. Uncertainty associated with incomplete knowledge of a phenomenon-and incomplete knowledge of the limits of one's knowledge-is referred to as epistemic uncertainty. Here, we illustrate the use of qualitative reasoning (QR) as a modeling approach that supports simulation despite pervasive epistemic uncertainty in a system. We develop a QR model of a simple plant-resource system to illustrate how six sources of epistemic uncertainty can be expressed, assessed, and managed. These include uncertainty about system structure, quantity vagueness, functional relationships, unknown or exogenous processes, simulation outcomes, and post-diction or explanation of outcomes. We show that QR provides a useful framework for expressing uncertainty due to inexact knowledge about parameter values and for exploring the consequences of different understandings of system structure. Furthermore, uncertainty in parameter values can be expressed and managed using different representations of and constraints on parameter quantity spaces. Compositional modeling supports the creation of alternative models representing different system structures. QR models allow the creation of a full envisionment of all possible outcomes given a set of causal processes, a particular system structure, and starting values. Finally, explicit representation of system structure and causality allows simulation results to be unambiguously explained. These features of QR support ecologists in making explicit their substantial qualitative knowledge about causes and effects in systems to produce models that give rise to insightful simulations of system dynamics.