Above-ground allometric biomass and BEF equations were developed in Eucalyptus nitens crops, in age sequence from 2 to 5 years and tree density between 2300 and 5600 ha(-1). All models were fitted for crown, stem and total above-ground biomass at tree and stand level and explained a high percentage of data variability (R-adj(2) >0.90). Biomass Expansion Factors (BEFs) were calculated for all categories and showed great variation, mainly for crown and total biomass. BEFs and stand-tree variable behaviour was analysed to develop BEF models to improve the predictions of constant BEF calculated here. Although all studied variables had significant relationships with BEF, dominant height showed the closest correlation with the crown and total biomass, the equations explaining 99% of biomass variability. Quadratic mean diameter, basal area and age were selected for the stem model. They explained more than 87% of the stem biomass variability. The comparison of the three approaches, biomass and BEF equations and constant BEFs, showed that biomass equations provided the most accurate predictions for stem and total components, followed by BEF equations. Constant BEFs proved the least accurate method for estimating biomass and only provided satisfactory results in relation to stem biomass. In contrast, for the crown component, BEF equations provided slightly more accuracy predictions than biomass equations. The best methodology for biomass production estimation depends on available resources and the level of required accuracy; however, our results suggest that constant BEF should be avoided whenever possible, at least for crown and total aerial biomass. (C) 2013 Elsevier Ltd. All rights reserved.