Cassava is an important tropical root crop for food security and national economies. In Ghana, the roots are used in popular local cuisines as well as in brewery, bakery, confectionery and plywood industries. A number of high-yielding and disease-resistant varieties are released and promoted to increase productivity and improve rural welfare. The study used a sequential mixed-method approach to identify, among drivers and impediments, the dissemination mechanism with highest impact on the adoption of improved cassava varieties (ICVs) and its intensity. The analyses helped estimate the impact of ICV adoption on productivity and households' livelihood, and to provide evidence of technological, managerial, and environmental gaps between adopters and non-adopters. Data were collected in 2014 from 608 randomly selected cassava-producing households in 14 communities in six districts of the Ashanti and Brong-Ahafo regions. Summary statistics reveal a 25 percent ICV adoption rate. Econometric analyses indicate significant and positive effects on the likelihood of households' ICV adoption for group members, the number of varieties planted, the number of livestock owned and information received mostly through innovation platforms (IPs). Impediments to ICV adoption include the location in the Ashanti region, household size, distances to the nearest tarred road and market, and grey-skin colour of ICVs. Results from propensity score matching and instrumental variable approaches indicate positive impacts of ICV adoption on cassava and whole-farm productivities and on per-capita annual crop income. Adopters appear to incur lower total annual per-capita expenditures and expenditures on food than non-adopters but spend more on children’s education. Bias-corrected stochastic output distance functions and stochastic metafrontier production functions showed strong evidence of technological, managerial, and environmental gaps between adopters and non-adopters in both cassava and whole-farm production. In both cases, adopters were found to operate on higher frontiers and to be more efficient than non-adopters. Adopters also appear to operate in a more favourable 2 production environment than non-adopters. The study provides strong evidence of inefficiency in cassava production for both ICV adopters and non-adopters. Findings imply that policy measures could be taken to increase the 25 percent ICV adoption rate through the establishment of IPs, focusing on households in Brong-Ahafo and those who are group members that integrate livestock-farming with cassava production. ICV adoption is expected to lead to increased productivity through technological change and enhanced efficiency. Moreover, the adoption of ICVs has the potential to increase crop incomes, food security and result in higher investment in children’s education, especially for female-headed households.