There is neither simple, nor universal relationship between mineral abundance and conflict. Pleased with the growing scholarly production on the local resource curse, we are concerned about how little we discuss the external validity of findings. Studies are sometimes taken as unconditional evidence of the resource curse (or its inexistence), which is a misunderstanding of a phenomenon that takes place under space and time-sensitive circumstances. To build our argument, we replicate and expand the period of analysis of Arellano-Yanguas (2011), a study of mining conflict in Peru that has become a point of reference in the new local resource curse literature. We show that the relationship between mining and conflict in Peru is complex. In particular, econometric findings on how fiscal windfalls impact on conflict will depend on variable definitions and period of analysis. The main goal of our research is to highlight the need of contextualizing quantitative evidence of resource-based local development that, alike qualitative research, provide results plausible of generalization yet not necessarily externally valid.