Thirty-nine different species of waste biomass materials that include woody or herbaceous resources as well as nut shells and juice pulps were used to develop empirical equations to predict the calorific value based on the proximate analysis results. Ten different linear/nonlinear equations that contain proximate analysis ingredients including or excluding the moisture content were tested by means of least-squares method to predict the HHV (higher heating value). Prediction performance of each equation was evaluated considering the experimental and the predicted values of HHV and the criteria of MAE (mean absolute error), AAE (average absolute error), and ABE (average bias error). It was concluded that the presence of moisture as a parameter improves the prediction performance of these equations. Also, the samples were classified into two subsets according to their fixed carbon (FC)/ash values and then the correlations were repeated for each subset. Both the full set of samples and the subsets showed a similar trend that the presence of moisture in equations enhances the prediction performance. Also, the FC content may be disregarded from the equation of the calorific value prediction when the FC/ash ratio is lower than a given value.