I discuss two classes of models that predict patterns of dispersal across heterogeneous environments; the source-sink and balanced dispersal models. Data collected during a long-term study of small mammals in an experimentally fragmented landscape support the predictions of the balanced dispersal model. In addition, a literature review indicates that for 28 species, few empirical studies have enough detailed demographic data to distinguish between assumptions of the two models and to test their predictions. The review indicates both models are poorly supported by empirical studies and many of the studies cited as evidence for source-sink dynamics cannot distinguish a sink habitat from a habitat with a low carrying capacity. I conclude that a continuum should exist between species vagility and the dispersal patterns seen in heterogeneous landscapes. Species with high vagility and the ability to assess habitat quality should tend towards balanced dispersal. In species with low vagility or where dispersal is controlled by abiotic factors (wind, or water currents), source-sink dynamics will be more common. Investigators must collect detailed demographic and movement data if they hope to test these models and further our understanding of the role of spatial dynamics in ecology.