Variation in the erodibility of rock units has long been recognized as an important determinant of landscape evolution but has been little studied in landscape evolution models. We use a modified version of the Channel-Hillslope Integrated Landscape Development (CHILD) model, which explicitly allows for variations in rock strength, to reveal and explore the remarkably rich, complex behavior induced by rock erodibility variations in even very simple geologic settings with invariant climate and tectonics. We study the importance of relative contrasts in erodibility between just two units, the order of these units (whether hard rocks overlie soft or soft rocks overlie hard) and the orientation of the contact between the two units. We emphasize the spatial and temporal evolution of erosion rates, which have important implications for basin analysis, detrital mineral records, and the interpretation of cosmogenic isotope concentrations in detrital samples. Results of the landscape evolution modeling indicate that the stratigraphic order of units in terms of erodibility, the gross orientation of the contact (i.e. dipping away or toward the outlet of the landscape) and the contact dip angle all have measurable effects on landscape evolution, including significant spatial and temporal variations in erosion rates. Steady-state denudation conditions are unlikely to develop in landscapes with significant contrasts in rock strength in horizontal to moderately tilted rock layers, at least at the scale of the entire landscape. Additionally, our results demonstrate that there is no general relation between rock erodibility and erosion rates in natural settings. Although rock erodibility directly controls the erosion rate constant in our models, it is not uncommon for higher erosion rates to occur in the harder, less erodible rock. Indeed erosion rates may be either greater or less than the rock uplift rate (invariant in time and space in our models) in both hard and soft rocks, depending on the local geology, topography, and the pattern of landscape evolution.