International regulatory representatives have proposed the performance-based test methodology for hydrogen fuel cell vehicle (HFCV) fuel system integrity certification in a new global technical regulation (GTR). For this test method, vehicle certification depends on system performance during barrier/rollover crash tests. The GTR proposal specifies that the test is failed if within 1 h post-crash, hydrogen leakage rates exceed 118 L/min or flammable mixtures develop within the passenger cabin or trunk. An analysis of the capabilities necessary to detect the second failure mode was performed through exploratory in-vehicle leakage tests at SRI International's Corral Hallow Experimental Site. Hydrogen concentrations were primarily derived from oxygen depletion sensor measurements, and were compared to directly measured concentrations from co-located hydrogen sensors. Close agreement between the two sensor technologies was observed. Since oxygen depletion measurements have the additional advantage that nonflammable gases can be used, helium was investigated as a surrogate due to its similar diffusion and jet spreading characteristics. The good agreement in overall dispersion trends for both gases highlights the flexibility of the indirect sensor method. While hydrogen mixture fractions strongly depended on release characteristics (e.g., rate, location, type), the results of an analytic examination indicated that pinhole leaks from moderate source pressures likely would produce unacceptably high in-vehicle hydrogen concentrations. The optimum sensor location for leak detection was determined to be high above the release point. Accordingly, sensor placement for crash tests involving vehicle rollovers must account for the final vehicle orientation. Copyright (C) 2011, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.