Ecological risk assessments are used by policy makers and regulatory agencies for balancing and comparing ecological risks associated with environmental hazards. An approach for regional—scale ecological risk assessment is described and demonstrated by modeling environmental risks associated with elevated ozone in a forested region. The demonstration illustrates (1) how a regional—scale risk assessment might be done, (2) the importance of spatial characteristics in considering regional—scale risk, and (3) the necessity of considering terrestrial and aquatic linkages. Generic problems often encountered when doing regional assessments, the foremost of which is the frequent lack of region—specific information and spatial data, are also highlighted. In the demonstration, two levels of elevated ozone and five different at—risk regional features are considered (forest cover, forest edge, forest interior, landscape pattern, and lake water quality). The mechanism for impacts on these features is ozone—induced stress in coniferous trees, patches of which can then be killed by bark beetle attacks. A stochastic spatial model of land—cover change is developed to evaluate the risks or probabilities of significant changes in the selected ecological features as a consequence of these ozone—triggered beetle attacks. Risk to regional water quality of lakes is evaluated by linking the land—cover output from the spatial stochastic model to an empirical water—quality model that is sensitive to land—cover changes within a lake's watershed. The risk analysis shows that those environmental features that are sensitive to the location of coniferous forest (such as forest edge) are at risk of a significant change due to ozone—induced conifer mortality even though overall coniferous forest cover is only slightly affected. The analysis also suggests a high probability of changes in regional water quality of lakes as a consequence of location—specific forest—cover change.