This paper presents the development of a linear programming model to determine the optimum fuel mix for greenhouse heating. The model is applied to two cases (a 7.5 hectare (ha) vegetable greenhouse and a 2 ha flower greenhouse) in British Columbia, which use a combination of wood pellets and wood residue to provide heat. The objective of the model is to minimize the annual total cost of combusting wood pellets and wood residue, while the heat demand is satisfied and all resources and emission limitations are not exceeded. The boiler system was enhanced with an electrostatic precipitator to limit the particulate matter emissions to acceptable levels. The results show that the optimal fuel mix for a 2 ha flower greenhouse is 641 tonne (t) of wood pellets and 381 t of wood residues, and it is 10 106 t of wood pellets and 3007 t of wood residues for a 7.5 ha vegetable greenhouse. The minimized annual total cost, including amortized capital cost, is estimated to be C$216030 y(-1) and C$1 611 866 y(-1) for a 2 ha flower greenhouse and a 7.5 ha vegetable greenhouse, respectively. A sensitivity analysis indicated that these optimal solutions would not change even when the variable costs (material cost and emission fee) of wood pellets decreased by 58% or the variable costs of wood residue increased by 150%. Copyright (C) 2008 John Wiley & Sons, Ltd.