Chlorine 36 was generated from nuclear tests in the 1950s and 1960s and has been used to identify fast flow paths at Yucca Mountain, the proposed repository for high-level nuclear waste <Fabryka-Martin et al., 1997, 1998>. Bomb pulse 36Cl, brought into the subsurface by infiltrating rainwater, presumably resides along fracture surfaces because of the extremely low rock matrix permeability. However, leaching a rock sample to extract this salt inevitably extracts pore water chloride (Cl) and rock matrix chloride, thereby making it difficult to obtain reproducible measurements or detect the specific bomb pulse signatures. Complexities introduced by these sources of older chloride include dilution of bomb pulse 36Cl/Cl ratios for samples from strata with a high Cl concentration, variations in measured ratios as a function of leaching time, rock chip size, and the differing effects of active leaching from those of passive leaching. This work provides both a conceptual model and a mathematical solution for the leaching processes and examines the role of sample leaching in the 36Cl studies of Yucca Mountain rocks. An analytical solution is derived for the diffusion of Cl and 36Cl in composite media (rock matrix and water) to accommodate variable diffusivity. This solution is subsequently used to develop a leaching model that takes into account bomb pulse signal, matrix pore water, and relatively hard to leach components (isolated fluid inclusion and mineral boundary salts). The model is then applied to samples from stratigraphic units at Yucca Mountain to obtain leachate concentrations from different setup methods (protocols), including duration, chip size, and gravitational settling of the water-rock mixture. The model results show that the probability of detecting a 36Cl/Cl bomb pulse signal is severely diminished under longer leaching times and smaller rock fragment sizes and that leaching times of 1 to 10 hours are most likely to be successful in detecting a bomb pulse signal. Bomb pulse 36Cl/Cl ratios are more likely to be observed if pore water Cl concentrations were initially low prior to the introduction of bomb pulse carrying water.