Different 13 new formulae have been developed for estimating the calorific values of 20 different biomass samples from their proximate analyses data. In order to obtain these formulae, 20 biomass samples from different types were analyzed to determine calorific values and proximate analyses. The measured net heating values of the biomass samples varied between 15.41 and 19.52 MJ/kg. All correlations were developed by means of least squares regression analyses. Regression coefficients of the correlations range from 0.829 to 0.898. Standard deviations of the heating values determined from 13 different correlations are between 0.4419 and 0.5280. © 2009 Elsevier Ltd. All rights reserved.