This paper uses multiple rounds of panel data to assess the distributional implications of the variability in agricultural productivity in Nigeria and Uganda. It uses both a conventional decomposition and a regression-based inequality decomposition approach to estimate the impact of climate-induced variability in agricultural productivity. To mitigate the endogeneity associated with unobserved time-invariant and time-variant household fixed effects, we use rainfall shocks as a proxy for estimating the exogenous variability in agricultural productivity that affects consumption. Results suggest that a 10% increase in the variability of agricultural productivity tends to decrease household consumption by 3.7 and 5.2% on average for Nigeria and Uganda, respectively. Controlling for other factors, variability in agricultural productivity contributed to between 25% and 43% of consumption inequality between 2010 and 2015 for Nigeria; and 16% and 31% of consumption inequality between 2009 and 2011 for Uganda. We also show that variability in agricultural productivity increases changes in consumption inequality over time.