Approximate profit functions are estimated using time-series, cross-sectional, county level data for 12 midwest states. Measures of climate variability are included in the profit functions. Simulated impacts of climate changes on profits are derived. Results show that inclusion of measures of climate variation are important for measuring the impact of changes in mean temperature and precipitation levels. Failure to account for the impact of differences in variability leads to an overestimate of damages. If global warming increases diurnal variation, such increases would have negative impacts on the profitability of midwest agriculture.