This paper re-evaluates practical methods of forecasting corn and soybean basis in the eastern Corn Belt. The accuracy of forecast methods differs over the course of the crop-marketing year. At harvest, historical moving average forecasts perform best. Post-harvest forecasts may be improved at short forecast horizons (<8'12 weeks ahead) by combining historical moving averages and recent basis levels. Results suggest that using 3-to-5-year moving average forecasts for corn basis and a 2- or 5-year moving average for soybean basis from harvest through April. The accuracy of these corn and soybean basis forecasts decreases markedly during the summer months.