This paper investigates the factor structure of commodity forward curve dynamics using data from pulp and oil markets. The data used is swap contract quotes, and forward curves are derived from this data using an optimization algorithm. A three factor model explains 89% of the price variation of the oil forward curves and 84% of the price variation of the pulp curves. The factor structure, especially in pulp forward curves, is more complex than found in many other studies, conducted mainly using interest rate data. Possible reasons for this phenomenon are discussed.