To assess non-destructive above-ground biomass measurements in comparison to destructive measurements in willow (Salix sp.) coppice plantations, individual plant weights of twelve 4-year old willow clones (Salix viminalis L., S. dasyclados Wimm.) were estimated by constructing allometric relationships between stem diameter and shoot dry weight. The plants were also measured with a destructive method, i.e. harvest, weighing, and determination of wood dry matter content. The allometric relationships between shoot dry weight and stem diameter at 55, 85 and 105 cm from the shoot base, were compared. Mean differences between the destructively and non-destructively measured plant weights were small and only significant for two of the clones. For nine of the clones, the non-destructive assessments deviated less than 3% from destructively measured weights. The maximum deviation was a 7% underestimation of the mean plant weight. However, the resulting differences between methods increased with increasing plant weight for six of the clones. For most clones, the best fit of the model relating shoot dry weight and stem diameter was found at 105 cm above shoot base (r(2)-adj greater than or equal to 0.987) but high model fits also were found at 55 cm(r(2)-adj greater than or equal to 0.973) and 85 cm(r(2)-adj greater than or equal to 0.961) above shoot base. Plant survival was determined to obtain production assessments per unit area. Standing woody biomass after 4 years ranged from 24.6 to 38.1 t dry matter ha(-1) and plant survival from 71% to 98%. We conclude that non-destructive methods are reliable for a range of clones in commercial practice, while improvement is needed for purposes in which detection of small growth differences is crucial. (C) 2003 Elsevier Ltd. All rights reserved.