Results/Conclusions: Results indicate that common assumptions about land cover trends (including disturbance and natural variability), ownership, land use history, prevailing dynamics of change, and the representation of relationships between economic sectors are often inadequately documented and inconsistent with empirical evidence. Also, the distinctions between drivers of initial land-use change at local scales and factors that influence subsequent management decisions are often ignored. Examples illustrate the need for more careful testing and validation of model simulation results. Results highlight specific knowledge gaps and recommendations are offered to improve the scientific basis, consistency and comparability of sustainability assessments (e.g. clear assumptions that build from empirical understanding, multi-disciplinary analysis, and reference data sets). Despite current uncertainties surrounding land-use change modeling, areas of consensus are identified for practices that can measurably improve sustainability relative to business-as-usual. The results also illustrate the costs and complexities of demonstrating “sustainability” of land-based production systems. Policy incentives are more likely to achieve desired results if goals are clearly defined, measurable, and complemented by systematic monitoring, and if benefits are perceived to outweigh costs.