Seasonal variations of the price premium for bread and feed wheat indicate opportunities to profitably adjust the grain marketing strategy of farmers that harvest (and store) both wheat types. In this article, we estimate the seasonal pattern of wheat price premiums on the German market using a vector error correction approach, which accounts for multivariate autoregressive conditional heteroscedasticity of the error terms. Our results indicate a significant downward trend for the premium during the marketing season, with the trend's magnitude depending on the average quality of harvested wheat. If farmers separately store both bread and feed wheat, they should tend to sell bread wheat before they sell feed wheat, particularly in years of low average wheat qualities. Further, the volatility of the price premium significantly decreases over the season, indicating higher risks at the beginning and lower risks towards the end. The volatility of the premium may present opportunities that require a reevaluation of the recommended temporal sequence of selling bread and feed wheat.