Briquetting of biomass can be done through various techniques. The present work describes the process of designing a taper die and its optimisation for use in a hydraulic machine. The application of an experimental design technique, and the later statistical analysis of the results is presented, applied to a laboratory hydraulic press densification process of lignocellulosic biomass. The most appropriate experiment type is determined for a first set of experiments; calculating, among other things: minimum number of tests to carry out to obtain binding conclusions, most influential factors, and search paths to improve fuel quality. Another experiment type is determined for a second set of experiments, taking account of the most influential factors [1, 3] (pressure, temperature and moisture content), and also the number of tests to carry out considering the improvement of density and friability. Finally, an approximation study of the best product allows conclusions to be reached on product behaviour beyond the experimental design range factors. (C) 2002 Published by Elsevier Science Ltd.