A study was conducted in the four counties the maize- wheat-teapotato and sugarcane-based farming system in North western Kenya to explore the variability among household characteristics and farm productivity. The aim of this work was to establish homogenous groups of crop-livestock mixed farming systems of Kenya. A two step approach was adopted for the study. The first was a rapid rural appraisal followed by a formal survey aimed at establishing farm types to facilitate detailed analysis of synergistic crop-livestock interaction systems. A random sample of 423 farmers was interviewed using a semi structured questionnaire. Descriptive and multinomial techniques were used in the analysis. Two classifications were utilized, the first step was establishing the criteria for classification and the second based on resource accessibility by farmers. The criteria used were the proportions of various crop and livestock enterprises and resource endowment. The process came up with eight farm types based mainly on the farm enterprise orientation, farm size, land productivity, cattle breeding, and farm by-products. Based on the formal survey three farm types were identified as intensive, semi intensive and extensive systems. The resource groups in all the counties were identified by: crop-livestock management, soil fertility management, food security and farm and off-farm income as important indicators of variability. However, all households were net food buyers, implying food insecurity. In addition, off-farm activities and off farm income were important livelihood survival strategies. Development planners and policy makers need to develop unique interventions targeting each specific group, since blanket policies are not appropriate in such a situation. Options such as optimizing livestock numbers to match available feed resources and improving feed availability through breeding and adoption of dual-purpose crop varieties with better digestibility coefficient, improving the cost-effectiveness of existing nutrition technologies (e.g. crop-by-products’), and bringing more land under fodder crops need to be explored. Since the farm sizes within the region had continued to decline, limiting the availability of on-farm livestock feed, there is need for policy instruments that can discourage land fragmentation