Governments at all levels expend resources to defend the environment from the adverse effects of human activity. Currently little is known about the processes and variables associated with these environmentally “defensive” expenditures. This paper attempts to fill this void by addressing the following research question: What factors induce municipal governments to expend scarce resources to defend against the adverse effectsof environmental change? The paper tackles this question with a combination of traditional statistical analysis and new environmental accounting techniques.Specifically, budgetary content analyses and secondary data are combined to develop a multiple regression model. The model uses the 1991 per capita defensive expenditures of 832 municipal governments as the response variable. These data are aggregated into 49 regions to ensure administrative commensurability. Model selection began with a wide array of predictor variablesthought to exert an influence on the defensive expenditures. The final model includes four predictor variables: provincial transfer payments for defensive expenditures per capita, property taxes per capita, median household income, and total pollution output per region. These four variables are significantly and positively associated with 73.5% of the variation in defensive expenditures (adjusted R2 = .735, p < .0001). Aninterpretation of the results suggests a need for a more place-sensitive approach to environmental accounting to ensure differences in environmental quality do not confound the analysis.