Climate-smart agriculture (CSA) is being widely promoted as a solution for food insecurity and climate change adaptation in food systems of sub-Saharan Africa, while simultaneously reducing the rate of greenhouse gas emissions. Governments throughout Africa are writing policies and programs to promote CSA practices despite uncertainty about the ability for practices to meet the triple CSA objectives of CSA. We conducted a systematic review of 175 peer-reviewed and grey literature studies, to gauge the impact of over seventy potential CSA practices on CSA outcomes in Tanzania and Uganda. Using a total of 6,342 observations, we found that practice impacts were highly context (i.e. farming system and location) specific. Nevertheless, practice effect across CSA outcomes generally agreed in direction. While our results suggest that CSA is indeed possible, lack of mitigation data precludes a more conclusive statement. Furthermore, the inclusion of potential adoption rates changes the potential of CSA practices to achieve benefits at scale. Given the uncertainty and variable impacts of practices across regions and outcomes, it is critical for decision makers to prioritize practices based on their desired outcomes and local context.