BACKGROUND: A reliable procedure for the determination of the volumetric oxygen transfer coefficient (K(L)a) in hydrocarbon-aqueous dispersions is critical to the assessment of oxygen transfer in hydrocarbon-based bioprocesses. The pressure step procedure (PSP) and gassing out procedure (GOP) for K(L)a determination were evaluated in n-C10-13 alkane-aqueous dispersions over agitation rates (600 to 1200 rpm) and alkane concentrations (up to 20%) typical of hydrocarbon-based bioprocesses. RESULTS: K(L)a from the procedures diverged progressively as agitation increased, with K(L)a underprediction attenuated at high alkane. At 1200 rpm and 5% alkane, K(L)a from the GOP was underpredicted by 49%. Compared with K(L)a from the PSP, K(L)a from the PSP without nitrogen transport was the same, but K(L)a from the PSP without response lag (PSP-lag) was underpredicted by approximately 50% at 1200 rpm. Moreover, K(L)a from the PSP-lag was lower than K(L)a from the GOP, except at 1200 rpm and alkane concentration up to 5%. CONCLUSION: The PSP was superior to the GOP in alkane-aqueous dispersions, particularly at extreme agitation and low alkane concentrations. This was attributable to the probe response lag and the non-uniformity of gas composition inherent in the GOP and independent of the rate of coincident nitrogen transport. (C) 2009 Society of Chemical Industry