Background Climate-Smart Agriculture (CSA) is fronted as a sustainable, transformative, and technologically innovative approach that increases agricultural productivity, incomes and enhances greenhouse gas mitigation. However, there is limited micro-level evidence on the effects of the adoption of CSA on food security despite intensified promotion efforts in Uganda. Methods A cross-sectional household survey among 165 respondents, undertaken in August-September 2020, was used to collect requisite data. Principal Component Analysis (PCA) with iteration and varimax rotation and Analysis of Variance (ANOVA) were used in characterizing CSA practices. An Ordered Logit Model was applied to identify the reported levels of CSA utility. Meanwhile, an endogenous switching regression was adopted to determine the effect of CSA adoption on household food security. Results Results showed that households used a combination of practices, including soil and water management, pasture management, livestock productivity and disease management. The PCA results revealed six major categories for the 16 most commonly used CSA practice combinations. The key factors that influenced the adoption of CSA practices among households included; access to climate information, total livestock units, ownership of non-livestock assets, and participation in off-farm activities. Results also revealed that the expected food consumption scores of adopters and non-adopters were 53.87 and 66.92 respectively. However, when adopters and non-adopters were compared, we found that the adopters of CSA practices would have had a significantly lower counterfactual food consumption score had they not adopted CSA. Conclusions While the adoption levels of CSA in this study is low, the counterfactual effects have shown that households that adopted CSA would have had a lower food consumption scores and therefore lower food security status had they not adopted CSA. We recommend CSA promotional efforts that give more attention to combined CSA practices and respond to local production constraints.