This paper applies a methodological framework for incorporating current period weather and long-term climate conditions into impact assessments. More specifically, the framework applies to non-experimental impact assessments that rely on ex post data collected from both households that were beneficiaries of the project (treated households) and those that did not benefit (control households). Here, the authors apply the methodological framework to an IFAD project that aimed to increase high quality coffee and the performance of coffee cooperatives in Rwanda as a case study. Results show that there is some evidence of biased treatment impacts when climatic variables are not included, but more importantly, show that coffee producers are highly vulnerable to weather shocks. To generate more climate-change relevant evidence more rapidly, there is ample opportunity to more fully exploit impact assessment datasets than is commonly done.