Biofortification, breeding staple food crops to be dense sources of essential micronutrients, is fast emerging as a strategy to fight micronutrient malnutrition. Large scale biofortification investments are being made in several developing countries, but until recently little rigorous evidence about the impact of these investments has been available. In this paper, we report findings from randomized impact evaluations conducted in both Mozambique and Uganda to study the impact of large-scale pilot projects conducted between 2006 and 2009 to introduce provitamin-A-rich orange-fleshed sweet potato (OFSP) as a strategy to reduce vitamin A deficiency. In both countries, projects randomly assigned interventions of different cost and intensity to distribute OFSP vines, train households to grow OFSP, and disseminate the health benefits of vitamin A. We compare the impact of the interventions within and across the two countries on OFSP adoption, knowledge about vitamin A, and dietary intake of vitamin A by children, and use causal mediation analysis (Imai et al. 2011) to examine the impact pathways on vitamin A consumption. After two years of intervention, in both countries the project led to OFSP adoption rates of 61-68 percent among project households, improved household knowledge about vitamin A, and nearly doubled average dietary intake of vitamin A, with no difference between the more and less intense intervention models. Evidence suggests that vine access played the most important role in explaining the impact on vitamin A consumption in both countries. Consequently, future programs can be designed to have similar impacts at even lower costs