This report is a review of the requirements and possibilities of developing a regional scale, long term model for the response of forests to atmospheric pollution. A general input-output structure of such models is delineated, and the input, output and potential model structures are considered. With the objective of specifying the input variables, potential causes of forest damage are classified according to their dynamic properties. Forest damage is further specified, so as to define the output variables. Theoretical and empirical models for the response of forests to air pollution are reviewed, as well as some empirical evidence applicable to the models. The possibilities of constructing scenarios of damage over Europe are assessed, and the present theoretical and practical restrictions are evaluated.