This study estimates the effects of a large discrete maize price increase on the welfare of a sample of rural Kenyan households. The usual first-order welfare approximation formula is extended to a second-order formula that allows for supply and demand responses to the price change. Results show that many rural households are not affected greatly by the price change, and there are about as many gainers as losers. However, these full sample results mask important differences across regions. Welfare gains generally take place in major production areas while losses are in areas where most households are net buyers of maize. Semiparametric methods are used to investigate the relationship between income and the size of the welfare effect, and poverty dominance techniques are applied to study the impacts of the maize price increase on rural poverty.