In the past, containment of hydraulic fracture height growth has been evaluated based on an assumption of rock formation layers with contrasting conditions of minimum horizontal stress, and to a lesser extent, Young's modulus, leak off rates, and fracture toughness between adjacent rock layers. Most recently, large-block hydraulic fracturing experiments in the laboratory, and observations of fracture propagation (natural or induced) in core, have provided evidence that the rock fabric plays a significant role in arresting fracture height growth and also in promoting fracture complexity. In addition, unconventional reservoirs are often over-pressured. And, as the pore pressure increases, the stress contrast tends to be reduced, and the role of rock fabric becomes dominant. In this paper, we investigate the effect of weak interfaces on fracture geometry and height containment by conducting hydraulic fracturing tests on large blocks from tight shale outcrops, under simulated effective stress conditions. We define rock fabric as the presence, orientation and distribution of bed boundaries, lithologic contacts, mineralized fractures, and other type of weak interfaces. This rock fabric creates discontinuities in the stress and strain fields and affects the way the rock deforms and fails. Continuous monitoring of acoustic emissions and using acoustic transmission during fracturing, allows understanding the process of fracture initiation and fracture interaction with the weak interfaces. Post-test CT x-ray scanning and detailed dissection and photographic imaging provide a good record of the fractures. In addition, these post fracture measurements allow comparing the fractures created with results from acoustic emissions localization. The experimental results clearly demonstrate the importance of rock fabric to understand and predict fracture complexity and fracture height containment.