The goal of landscape ecology is to understand the relationships between landscape pattern and ecological process; the role of human impacts and other forces of landscape change on these pattern–process relationships; and the principles required to make informed decisions in natural resource management. The interdisciplinary nature of the science of landscape ecology has produced a diverse set of models varying in purpose, methods, and complexity. In spite of this diversity, landscape models have a number of similar attributes that provide an organizational framework for our definition: (1) at their most basic level, they provide maps of one or more attributes of landscape heterogeneity (pattern); (2) they consider the influences of landscape spatial heterogeneity on ecological interactions and fluxes (pattern on process); and (3) they illustrate how changes in landscape pattern can be generated by spatially explicit ecological processes such as disturbance (process on pattern). No single paradigm exists that adequately defines how landscape models should be designed and tested. Nevertheless, the requirement that landscape models consider the relationships between spatial heterogeneity and ecological processes creates a number of issues that require special attention if useful and reliable results are to be obtained. Three such issues are landscape history, scale of analysis, and empirical requirements for model evaluation.