The past decade has seen renewed interest in the use of mixed-method approaches across the social sciences and in the field of impact assessment. This body of work has focused on questions of method with insufficient attention devoted to foundational issues. The objective of the present article is to bare the foundational concepts that guided a mixed-method impact assessment of the Hunger Eradication and Poverty Reduction (HEPR) programme in Vietnam. Specifically, it shows how the applied methods used in the HEPR study rested on foundational differences concerning: conceptions of causation and models of causal inferences (probabilities versus mechanisms); analytical focus (outcomes versus processes) and external validity (empirical generalisation versus statistical inference); and constituents of ‘objective’ knowledge (intersubjective observables versus perceptual data).