Context Landscape structure influences host-parasite-pathogen dynamics at multiple scales in space and time. Landscape epidemiology, which connects disease ecol-ogy and landscape ecology, is still an emerging field. Objective We argue that landscape epidemiology must move beyond simply studying the influence of landscape configuration and composition on epidemiological processes and towards a more comparative, systems approach that better incorporates social-ecological complexity. Methods We illustrate our argument with a detailed review, based on a single conceptual systems model, of geographic variation in drivers of avian influenza in Western Europe, Southeast Asia, and Southern Africa. Results Our three study regions are similar in some ways but quite different in others. The same underlying mechanisms apply in all cases, but differences in the attributes of key components and linkages (most notably avian diversity, the abiotic environment, land use and land cover, and food production systems) create significant differences in avian influenza virus prevalence and human risk between regions. Conclusions Landscape approaches can connect local-and continental-scale elements of epidemiology. Adopting a landscape-focused systems per-spective on the problem facilitates the identification of the most important commonalities and differences, guiding both science and policy, and helps to identify elements of the problem on which further research is needed. More generally, our review demonstrates the importance of socialecological interactions and comparative approaches for landscape epidemiology.