This study investigates the effects of proximate, ultimate and elemental analysis for Afghan coal samples on Hardgrove grindability index (HGI), Gross calorific value (GCV), and Ash fusion temperatures (AFTs) by using multivariable regression (MR) and Adaptive neuro-fuzzy inference system (ANFIS) to increase information about the properties of the Afghan coal. Statistical modeling (MR, and ANFIS) indicated that coal parameters (HGI, GCV, AFTs) can be predicted with high accuracy, where GCV, AFTs, and HGI were estimated by R-2 = 0.99, 0.95, and 0.94, respectively. The small difference between the estimated parameters and their actual values shows that these accurate results can be also applied to estimate coal properties in other coal resources of Afghanistan. (C) 2012 Elsevier B.V. All rights reserved.