This study aims at describing the baseline situation of crop-livestock systems in the semi- arid regions of Zimbabwe. Using quantitative household surveys data and stakeholder workshops, it captures livelihood strategies, community visions, crop and livestock production and marketing, perceptions of risk and uncertainties and farm household typologies. The present data set captures quantitative household data from two districts of Gwanda and Nkayi. At each site multi-stage sampling approach was used. A total of four wards in each district were purposively selected, to include wards close to and far from market centres and in each ward three villages were randomly selected. Household village population across the two districts ranged from 200-300 households. Using random sampling techniques 25-30 households were selected in each village. The sample size (n) for Gwanda is 350 and for Nkayi is331. The overall goal of the project is to identify, test and prove ways to increase agricultural production, improve household food security, alleviate poverty and thereby reduce food aid dependency in rural Zimbabwe through better integrated crop and livestock production and market participation. The project integrates a participatory technology development framework , including baseline diagnostics, stakeholder workshop, systems simulation modeling, technology screening, on farm trials and demonstrations, with value chain analyses that feed into multi-stakeholder platforms for knowledge exchange thereby linking to effective impact pathways.Two districts of Goromonzi and Murehwa are in the semi-humid region and the other two districts of Gwanda and Nkayi are in the semi-arid region of the country.