<p>A major challenge for countries dealing with conflict and instability is promoting the use of climate-smart farming technologies and practices. In this meta-analysis, we assess determinants of farmers’ adoption decisions for such innovations. We employ advanced machine learning-aided literature selection and review 112 papers selected from over 42,000 published papers covering countries in fragile and conflict settings. We categorized the technologies into five technology groups, including soil health, erosion management, mechanization, input use and risk reduction technologies and extract 1374 coefficients. Univariate and multivariate partial correlation coefficient analysis suggests that factors such as training, access to information, subsidies, and past experiences of using technologies predict technology adoption. However, there are significant differences across technology groups and most especially, a very low coverage of risk-reduction technologies such as insurance is recorded.</p>