Prior to the 2000 planting season, some industry observers predicted acreage of genetically modified crops would decline dramatically. However, actual 2000 plantings presented a puzzle. Farmers reduced their acreage of genetically modified corn, but concurrently increased their acreage of genetically modified soybeans. We demonstrate that it may be theoretically optimal for risk-averse farmers to reduce their corn acreage but not their soybean acreage. However, past experience, attitudes, and farm size explained planting decisions to a larger degree than did risk preferences.