Coal continues to dominate India&rsquos energy sector, with underground mining gaining increasing importance as open-castable reserves depleting or environmentally restricted. Development of coal seam by Bord and Pillar method, followed by systematic depillaring with or without stowing, remains the most widely practiced extraction technique due to its operational flexibility. However, the safety and productivity of this method depend critically on the stability of coal pillars, particularly during depillaring when load redistribution is severe. Conventional empirical approaches, such as those of Salamon&ndashMunro (1967) and Sheorey (1992), although widely adopted, assume homogeneous and isotropic strata, and therefore overlook the role of weak beds or dirt bands that are commonly present in Indian coal seams. These thin layers of shale, claystone, or carbonaceous material significantly compromise confinement, modify stress redistribution, and accelerate pillar failure, often leading to unsafe or overly conservative designs. The present research addresses this limitation by systematically investigating the influence of weak beds on coal pillar strength and proposing a refined design framework suitable for stratified formations. The study integrates multi-mine field investigations, empirical assessment, and three-dimensional numerical simulations using FLAC3D software. Geological and mining data were collected from selected SCCL and SECL collieries, where depillaring is practiced under weak and laminated conditions. Empirical analysis revealed that conventional formulas consistently overestimated safety factors in panels affected by weak beds. Parametric numerical analyses evaluated the effects of weak bed position, thickness, and continuity, along with depth of cover, pillar geometry, and rock mass properties. The results indicated that roof-interface weak beds produced the most severe reductions in strength up to 20&ndash25% under deep cover&mdashprimarily due to gallery widening and development of non-elastic zones in pillar ribs. Mid-pillar partings induced shear and delamination, while floor-embedded beds triggered floor heave and confinement loss. Even very thin layers (0.2&ndash0.5 m) caused significant weakening, and multilayer conditions showed cumulative effects leading to progressive through-thickness shear failure. To reconcile these discrepancies, correction factors typically ranging between 1.12 for isolated weak beds and 1.18 (or higher, up to ~1.30 under multilayer or deep cover conditions) were derived to quantify the weakening influence of depth, weak bed configuration, and rock mass parameters. These were incorporated into Sheorey&rsquos empirical formulation to propose a modified pillar strength equation. Reinforcement measures, including side bolting, were also numerically evaluated and found effective in restoring confinement and control spalling, thus improving safety margins. This research bridges the gap between empirical simplicity and mechanistic understanding by offering numerical models calibrated framework for pillar design. The findings enhance the reliability of stability assessment, optimize coal recovery, and provide scientific input for updating statutory design practices in India. Beyond its immediate relevance, the methodology is generalizable to stratified coalfields worldwide, contributing to safer and more efficient underground mining practices.