The analysis is based on an integrated climatic, agronomic and economic model of grain production in the midwestern region of the U.S. The model represents a typical farm and the objective is to select an optimum combination of cultivars for four crops (maize, soybean, wheat and sorghum) conditionally on climate, soil characteristics and the prices of the four crops. The first part of the analysis uses a complete two-way factorial design to estimate the effects of five climatic and agronomic variables and the four prices on yields, total production and net return. With this design, variability can be allocated uniquely to individual explanatory variables. As expected, prices have little effect on yields. For total production and net return, temperature and precipitation are the most important variables, but prices and two-way interactions among the variables are also statistically significant. The ranges of values for temperature and precipitation used in the factorial design are extended in the second part of the analysis. Response surfaces for yields, total production and net return are estimated from these new data. The fitted surfaces are used to evaluate the effects of projections of temperature and precipitation, resulting from global warming, on yields, total production and net return. Two of the three projections imply modest increases in yields, total production and net return and one shows slight declines, particularly for the yield of maize. The differences in these results reflect primarily differences in the projected changes in precipitation because the projected increases of temperature are similar in all scenarios. Furthermore, relatively small reductions in precipitation from the projected values have relatively large adverse effects on grain production.