This study uses a randomized saturation experiment and influential technology promoters to test strategies to promote diffusion of two highly nutritious agricultural crop technologies in Uganda to measure which approaches are most cost-effective at achieving high adoption rates. The crops are conventionally bred varieties of vitamin-A-rich orange sweet potato (OSP) and high-iron biofortified beans (HIB), distributed as an intervention to reduce vitamin A deficiency and anemia. The experiment included four treatment arms: three levels of randomized saturation of households with the crop technologies (control-0%, low-20%, and high-50%) and a treatment in which opinion leaders in farming and health identified through an election were invited to promote the technologies. Results show that being assigned to treatment in either the low or high saturation substantially increased the average probability of adopting the crops over the five seasons of the project and increased spillovers to neighboring households by 16-19%. There was no difference in spillover effects between the high and low saturation treatments. In addition, the low saturation treatment for high iron beans had a reinforcing effect on adoption by neighboring farmer group members who were also given access to the treatment, increasing their probability of adopting the crop by 14.9 percent in the last season of the project. The opinion leader treatment led to no more diffusion of either technology than in the control group.