We modelled the kinetics of H2 flux during gas uptake and elimination in conscious pigs exposed to hyperbaric H2. The model used a physiological description of gas flux fitted to the observed decompression sickness (DCS) incidence in two groups of pigs: untreated controls, and animals that had received intestinal injections of H2-metabolizing microbes that biochemically eliminated some of the H2 stored in the pigs' tissues. To analyse H2 flux during gas uptake, animals were compressed in a dry chamber to 24 atm (ca 88% H2, 9% He, 2% O2, 1% N2) for 30-1440 min and decompressed at 0.9 atm min(-1) (n = 70). To analyse H2 flux during gas elimination, animals were compressed to 24 atm for 3 h and decompressed at 0.45-1.8 atm min(-1) (n = 58). Animals were closely monitored for 1 h post-decompression for signs of DCS. Probabilistic modelling was used to estimate that the exponential time constant during H2 uptake (tau(in)) and H2 elimination (tau(out)) were 79 +/- 25 min and 0.76 +/- 0.14 min, respectively. Thus, the gas kinetics affecting DCS risk appeared to be substantially faster for elimination than uptake, which is contrary to customary assumptions of gas uptake and elimination kinetic symmetry. We discuss the possible reasons for this asymmetry, and why absolute values of H2 kinetics cannot be obtained with this approach.