Instructional technologies critically depend on systematic design, and learning hierarchies are a commonly advocated tool for designing instructional sequences. But hierarchies routinely allow numerous sequences and choosing an optimal sequence remains an unsolved problem. This study explores a simulation-based approach to modeling learning hierarchies and the sequencing problem with a special focus on mastery learning over time. Computer simulations of eight hierarchies from the literature address two general questions about hierarchies and sequences: Do different valid instructional sequences for a given hierarchy lead to significant differences in simulated learning outcomes?; and How do hierarchy structures and student ability influence learning? Two findings emerge from the simulations: equally valid instructional sequences can produce widely divergent learning outcomes, and student ability can interact with sequence and hierarchy structure to disadvantage lower-ability learners. The paper concludes with observations on the role of simulations in the analysis and validation of instructional design and large-scale instructional systems.