This dissertation investigates the environment, activities and outcomes of three economic development programs delivered by the Government of Canada to communities of rural, northern Manitoba having substantial Aboriginal populations. These programs, delivered by different sections of essentially one evolving agency, operated over a 19 year period from 1971 to 1989. The research contains both exploratory and quasi-experimental components. Government and client socioeconomic environments are described qualitatively and quantitatively. Qualitative data collected through review of administrative files reveal processes of program design, operational patterns and change. Qualitative and quantitative data from nearly 1,600 applications for business financing are used to generate descriptive and analytical statistics concerning characteristics of applicants; the project intent of applicants; program response processes, decisions, and outputs; and project outcomes. Activity flows and attrition rates are explored within a causal systems model. Measures of applicant capacity, and outcome effectiveness and efficiency are applied to project data. Strengths, weaknesses and crucial tradeoffs in program design, given pressures and constraints imposed by the programs' environment, are uncovered. Project and program activity characteristics associated with higher business and employment payoffs are differentiated from project and program activity characteristics associated with lower business and employment payoffs. Points-of-interest and propositions are formulated from literature in the fields of economic development, public policy, and organizational structure and operation. Study findings are brought to bear on the not-testable, points-of-interest. Propositions are tested as formal hypotheses against descriptive and analytical statistics.