The objective of the study is rst to identify economic factors that in uence two speci c classes of wheat: hard red winter (HRW) and soft white (SWW) wheat, and develop models to improve the forecast performance of basis in Washington State. Earlier work has investigated basis behavior of some other classes of wheat, but none has examined soft white wheat. This class is unique because there is no direct futures contract-it is usually priced o the soft red wheat futures contract. The models we estimate include: 1) a simple moving average model to serve as a benchmark, 2) an econometric fundamental model, 3) an ARMA time series model, and 4) an ARMAX hybrid model. The econometric fundamental and ARMAX models include supply/demand factors suggested by economic theory and literature. We estimate all the models and then compare their forecast performance. Based on empirical results, we nd the best HRW model at the 4-month and 11-month forecast horizons is the econometric fundamental model, at the 5-month, is an ARMA(3,0,0) model, and the best model for the rest of the forecasts is an ARMAX (3,0,0). For SWW, the econometric fundamental model is the best overall. In addition, the ARMAX models perform better than the ARMA models in most cases, except SWW in Odessa, WA.