Shale resources have distinctive characteristics compared to conventional reservoirs, including microsized pores (IUPAC definition), ultralow permeability, several gas storage mechanisms, and complex fluid flow behavior. Prediction of productivity and deliverability of shale systems requires knowledge about in situ porosity and permeability. In this study, we evaluate pore and hydraulic connectivity of matrix for Barnett and Haynesville shale plays based on mercury injection capillary pressure (MICP) data and percolation theory. Using MICP porosity values measured at the laboratory for different sample size, accessible porosity and permeability for Barnett and Haynesville shale samples are reported. Next, pore and hydraulic connectivity for both Barnett and Haynesville samples are evaluated based on percolation theory. Moreover, permeability values that have been calculated based on MICP data are used to estimate the average coordination number as a function of sample size. Our results Indicate that accessible porosity and matrix permeability decreases with increasing sample size, which predicts lower connectivity for shale matrix in large scale. Consistent with percolation theory, results suggest that accessible porosity decreases with increasing sample size, following a power law function. Furthermore, results show that sample size has a significant impact on the estimated coordination number; this is expected, because interconnected porosity is a strong function of the average coordination number. The main contribution of this work is the evaluation of accessible porosity and pore connectivity for different sample sizes from two shale plays. The new insight about scale -dependent pore connectivity and interconnected porosity may lead to improved predictions of production performance and project economics.