Bubble populations and gas transfer velocities were measured in cleaned and surfactant-influenced freshwater and seawater. A nonlinear fitting technique was used to partition the total gas transfer velocity for a gas in each water type into a turbulence- and bubble-mediated fraction. This showed that the bubble-mediated transfer fraction was larger in cleaned freshwater than in cleaned seawater and that the difference was a function of diffusivity and solubility. This was explained by the fact that the bubble measurements showed that bubble plumes in cleaned freshwater had a higher concentration of large bubbles and a lower concentration of small bubbles than the plumes in cleaned seawater. The differences between the behavior of the bubble-mediated gas flux in cleaned freshwater and cleaned seawater show that caution should be used when intercomparing laboratory results from measurements made in different media. These differences also will make parameterizations of bubble-mediated gas exchange developed using freshwater laboratory data difficult to apply directly to oceanic conditions. It was found that adding a surfactant to seawater had minimal impact on the concentration of bubbles in the plumes. Because surfactants decrease the gas flux to the individual bubbles, the similarity in bubble population meant that the addition of surfactant to seawater decreased the bubble-mediated gas flux compared to the flux in cleaned seawater. In contrast, the addition of a surfactant to freshwater increased the concentration of bubbles by over an order of magnitude. This increase in bubble population was large enough to offset the decrease in the flux to the individual bubbles so that the net bubble-mediated gas flux in freshwater increased when surfactant was added. This difference in behavior of the bubble population and bubble-mediated transfer velocity between surfactant-influenced and cleaned waters further complicates interrelating laboratory measurements and applying laboratory results to the ocean.