Porosity is considered one of the most important rock physics parameter when evaluating a gas shale reservoir's production potential. The measurement of total porosity by the gas injection porosimetry (GIP) method in these low permeability rocks with complex mineralogy has usually proven to be challenging. It is not rare for results to vary from different laboratories but the reasons behind these inconsistencies are not fully understood. These differences are commonly attributed to apparatus (different performances of porosimeter), core cleaning (lacking of cleaning effect evaluation criteria and arbitrary drying) and measurement (insufficient gas injection pressure and inconsistent crushing level). In this study, systematic experimental investigations are conducted to study the impact of these factors in determining gas shale's total porosity using the GIP method. We propose a modified porosimeter design to meet the particular needs of gas shales. Helium is used to minimize adsorption via the molecular sieve effect that can occur during gas shale porosity measurements. ESH (extended slow heating) pyrolysis and TGA (thermal gravimetric analysis, 200 degrees C) techniques are also adapted to evaluate core cleaning effects ensuring that only integrated matrix is left in the sample. The experimental results indicate that the ESH pyrolysis technique produces more reasonable results than the Rock-Eval II pyrolysis in free hydrocarbon cleaning effect and solid organic matter integrity evaluation when experimenting on gas shale and other organic-rich rocks. Besides, the effects of gas injection pressure and particle size on porosities measured with plug samples and crushed samples are studied. Results show that the gas injection pressures play an important role in both crushed and uncrushed samples' porosity measurements. It is proven through this series of experiments that both the plug and GIP methods can effectively measure the total porosity of gas shales. The plug sample requires a higher minimum injection pressure and more total diffusion time in the total porosity estimation with respect to crushed sample, because these two parameters are greatly decreased by crushing. Note that the lower crushing level should be controlled not to alter the integrity of grain size composition. The porosity resulted by GIP method is interpreted as total gas accessible porosity. (C) 2016 Elsevier Ltd. All rights reserved.