Given that regional ecosystems are unique and thus may differ considerably in their normal ranges of primary and secondary productivity, species composition, diversity, and nutrient cycling, and given that each system is exposed to unique combinations of stresses, it might be expected that patterns of response to stresses will be highly variable and unpredictable. Therefore, it is surprising to discover remarkable similarities in the response of ecosystems to stress (Odum 1985, Rapport et al., 1985; Rapport and Regier, 1995). Stressed ecosystems are characterized by a "distress syndrome" (Rapport et al., 1985) that is indicated not only by reduced biodiversity and altered primary and secondary productivity but also by increased disease prevalence, reduced efficiency of nutrient cycling, increased dominance of exotic species, and increased dominance by smaller, shorter-lived opportunistic species. These signs have been well documented in a number of studies of both terrestrial and aquatic systems (Hilden and Rapport, 1993; Rapport et al., 1995; Whitford, 1995; Epstein and Rapport, 1996; Wichert and Rapport, 1998). How might this distress syndrome pattern be explained? By what mechanisms do stressed ecosystems become degraded? Why has it proven so difficult to rehabilitate stressed ecosystems, even after the initial stresses have been reduced or removed altogether? In this article, we address these and related questions by an empirical examination of three very different regional ecosystems, each of which has had a long history of exposure to multiple anthropogenic and natural stresses.