This paper investigates the effects of adopting drought tolerant maize varieties (DTMVs) on farm productivity and risk exposure using a moment based approach on households growing maize in Zambia. First, second and third moments of farm production were used in estimations. The study applied an endogenous switching regression model that controls for both observed and unobserved sources of heterogeneity between adopters and non-adopters. The study revealed that the adoption of DTMVs increases maize yield, reduces yield variability and exposure to downside risk significantly. The adoption of DTMVs increased maize yield by 8% while reducing yield variance and the risk of crop failure by 35% and 27%, respectively. These results underscore the need for concerted efforts to scale-out the production of DTMVs for both maize productivity enhancement as well as for risk mitigation against climatic shocks.