A methodology for rock classification is presented that considers lithology and reservoir parameters on the basis of a combined fractal analysis of well logs and mercury porosimetry results with nuclear magnetic resonance outcomes. A sandy-shaly thinly-bedded Miocene gas bearing formation in the Carpathian Foredeep is investigated. Fractal correlation dimensions D2 calculated for standard logs are used to distinguish sandstone as the most homogeneous lithological group with the highest porosity. The fractal analysis also confirmed observations of gas accumulations in sandstone, shaly sandstone and sandy claystone as rocks of high porosity. The results of two laboratory methods are combined to improve reservoir properties assessment and evaluate movable media in pore space; this technique was based on the similarity of mercury porosimetry results plotted as cumulative intrusion volume versus pressure or pore diameter and also the curves of cumulative porosity and transverse relaxation time distributions using nuclear magnetic resonance. Close values of porosity from logs, recorded in situ and in laboratory measurements, provide the link between fractal analysis and porosimetry and nuclear magnetic resonance measurements.