Since the late 1990s, U.S. production of corn ethanol has risen rapidly. In response to high demand, driven in part by rising ethanol production, corn prices and corn production surged in 2007 when corn plantings reached their highest level since 1944. To increase corn acreage, farmers shifted land to corn from other crops or, possibly, returned uncultivated land (e.g., cropland pasture, CRP land) to corn production. Even before 2007, however, "islands" of relatively high corn prices formed around ethanol plants in the Midwest. Price impacts were usually concentrated around an ethanol plant and ranged between 4.6 cents and 19.6 cents, with an average price increase of 12.5 cents at the plant site. Prices were also affected up to an estimated 68 miles from the plant (McNew and Griffith, 2005). Did these price island effects induce producers to shift to crop rotations that include more corn and/or bring in uncultivated land to corn production? If localized changes did occur in the years before 2007, they may persist into the future even though corn prices have declined absolutely and in relation to prices for soybeans and other crop commodities. The question is important because continuous corn, corn-intensive crop rotations, and shifting pasture or hayland into corn, can adversely affect the environment. This paper develops a discrete choice model that incorporates price island effects, local ethanol capacity, and broader land use change to understand the effect of ethanol-driven price islands on corn acreage, corn rotations, and general land use for cultivated crops. The primary data set in estimating the econometric model is the 2005 Corn Agricultural Resource Management Study (ARMS) survey, collected by the Economic Research Service and the National Agricultural Statistics Service. Because ARMS phase II (field-level) data is geo-referenced, spatially explicit data on corn and soybean prices are linked, along with the proximity of farms to ethanol plants and a soil productivity index. The ARMS Corn data is drawn from the traditional Corn Belt, along with some outlier states including North Carolina and North Dakota. A nested multinomial logit model (NML) is used to estimate producer crop mix response to local corn price islands, land quality, and other farm and location-specific factors. The NML model allows us to account for a range of crop production and land use options which vary in terms of similarity and, therefore, substitutability. At the highest level of the nested model, the farmer decides if he will cultivate his land or leave it uncultivated. If he chooses to cultivate his land, he needs to decide what crop to plant, for example, a corn-soybean rotation, wheat, or some “other” crop. The farmer will choose the crop and rotation pattern that will provide him with the greatest return, given prices and inputs. In this paper, the NML estimates the probability that a farmer will choose corn and soybeans (conditional on the choice of corn or soybeans) at the lower level and the choice among corn or soybeans, wheat, and “other” crops at the upper level, where the probability is a function of a corn/soybean price ratio; ethanol capacity index; livestock indicator variable; irrigation; soil quality and protected land statuses (highly erodible land); and household and farm characteristics. Because parameters cannot be directly interpreted in this model, the marginal effects and elasticities are examined. The authors find that in both levels of estimation, soil productivity and livestock value influence a farmer’s decision to plant corn or soybeans, wheat or some other crop. The estimation of our lower level confirms that local prices have a strong influence on whether a farmer will choose to plant corn or soybeans, while our upper level estimation may suggest that an increase in local ethanol capacity will encourage farmers to plant corn or soybean relative to both wheat and “other”. Information on the influence of "price islands" on farm behavior, including farm crop and rotation patterns and individual farmer land use decisions, could have environmental and other implications. This work could be extended by linking land use change to nutrient runoff and loads in water, possible soil erosion, and other environmental impacts from continuous corn rotations.