The value and volumes of industrial stump fuel supply are increasing for energy production. Accurate estimates of aboveground and belowground biomass of trees are important when estimating the potential of stumps as a bioenergy source. In this study two stump biomass equations were adapted and tested using them as calibrated stump biomass models computed as the cumulative sum by a local stand. In addition, variables derived from stem measurements of the forest harvester data were examined to predict stump biomass of a stand by applying regression analysis. The true stump yield (dry weight) was used as the reference data in the study. Both biomass models performed well (adjusted R(2 similar to)0.84) and no advance was found in using other stem dimensions as independent variables in the model. The stand-level model can be used in innovative stump biomass prediction tools for increasing efficiency of energy wood procurement planning to stands within a certain area. In practice, wood procurement managers would need to adapt developed system and decide whether the degree of accuracy/precision provided by the models is acceptable in their local stand harvesting conditions. (C) 2010 Elsevier Ltd. All rights reserved.