AbstractThis paper describes two research programs that are aimed at developing ecologically-based landscape planning models. The first of these models was developed as part of a more comprehensive landscape planning research program at the University of Massachusetts. The second is an on-going program in forest land planning at the University of Vermont. In addition to an emphasis on ecological values, both projects make extensive use of computerized geographic information systems (GIS) for data storage, analysis and display. The emphasis has been to develop a synthesis of planning procedures, ecological theories and quantitative techniques. Both research efforts were guided by a three-part methodological framework adapted from the work of Eugene Odum. The Massachusetts work generated a two-part classification system. The first part of the classification uses the statistical technique of discriminant analysis to place land uses into five groups, each having similar ecological characteristics. The second part of the classification is based on the physical and environmental characteristics of a site. The biological and denudational potentials of a site are determined from soil characteristics and slope. A comparison of the two classification schemes was made to determine ecological compatibility. The procedures were applied to two communities in Western Massachusetts. The Vermont study was aimed at integrating ecological values into forest land management and planning processes. The research utilized soils and topographic information in the development of an ecologically based forest land suitability model. Forest site index and soil erosion potential were determined using data stored in a computerized geographic information system. Forest site index estimates were made by utilizing an existing, well documented soil-site index regression equation applicable to the Green Mountains of Vermont. Soil erosion potential estimates were made by combining a soil erodability factor (K) with slope. The site index and soil erosion estimates were combined to produce land suitability classes for resource protection, forest management, multiple-use and trade-off. The procedure was applied in the Mad River Valley in Central Vermont. Both the Massachusetts and the Vermont studies are seen as significant contributions toward the integration of ecological factors into processes for landscape planning and resource management.