Non-destructive biomass estimation methods for Salix which were developed for research purposes in Swedish energy-forestry are too time-consuming and expensive to be applied in commercial field-situations. Existing allometric equations (relating shoot weight to shoot diameter) may be used to exclude the destructive, time-consuming phase in biomass-estimation procedures. These relations were shown to be specific for species, age, clone and sire, implying that generalized equations may lead to considerable errors for biomass estimations of particular stands. However, generalized equations were identified which, in general, caused deviations of less than 10% compared to the outcome of site-specific equations. Differences between estimated standing-biomass (> 5 cm above ground level) and biomass harvested at higher levels by commercial equipment can be corrected for and are species specific (-0.6 and -0.8% cm(-1) for Salix viminalis and S. dasyclados, respectively). While performing a random sampling of stools for non-destructive estimation field workers tend to sample around an 'ocular mean', which leads to an over-estimation of standing biomass.