||To address increasing threats of climate change and declining soil fertility, a cross regional agronomic study was conducted in southern Africa, covering 19 onfarm communities in contrasting agroecologies ranging from around 500 mm to more than 1800mm of rainfall. The 19 onfarm communities have been under longterm research by CIMMYT and national partners in Malawi, Zambia and Zimbabwe. The aim of this study was to assess how different climatesmart agriculture technologies perform under a variable climate and to pilot new doubledup legume systems of groundnuts with pigeonpea in 6 selected target communities. Yield results from all the sites showed primarily positive response of CSA as compared to the conventional control practices, although, these differences were not always significant. In 11 out of 19 communities there was a significant maize yield benefit recorded when CSA was practiced. In 9 out of 13 communities with full maizelegume rotation, the legume yield under different CSA practices had significantly greater yields than the conventional control treatment. Maize yield benefits of close to 100% were recorded in some sites, which shows great potential of these systems to withstand climate variability and change. Although these results were mostly positive towards CSA, the data from the 6 pilots were incomplete as the pigeonpea yields from the doubledup systems were not collected in full. Farmers planted the pigeonpeas too late, had problems with beetles that attacked the pots, had challenges with harvesting, which all contributed to inconclusive result. The adaptive capacity of CSA practices is widely acknowledged. However, the mitigation benefits is unclear and often leads to wrong statements about its potential. We therefore conducted a cross regional soil carbon study to assess how much soil carbon is being stored in the different systems. Soil carbon data from the sites was mostly the same between treatments and in some there was a positive response towards CSA. This was mainly in manual systems of Eastern Zambia and in some sites in southern Zimbabwe. In all other sites the soil carbon gain in onfarm systems was small. There are many reasons that could be responsible for the lack of carbon increase: grazing of crop residues by cattle, burning of residues by mice hunters, bush fires, and the long dry season that might have reversed the positive effect of soil carbon gain expected from CSA. Also in sites of Malawi, all cropping systems (even the conventional treatment) are in full rotation with legumes, and crop residues are often buried in the local conventional practice (e.g. the ridge and furrow system) which might have reduced this gain. Finally, the short duration of this study did not allow for a proper assessment of carbon sequestration or resilience in the 6 pilot trials and we would recommend that this project should be continued for another cropping season to better assess the new diversification elements that have been introduced by the project in the pilots.