The emission of biogenic gases from large rivers can be an important component of regional greenhouse gas budgets. However, emission rate estimates are often poorly constrained due to uncertainties in the air-water gas exchange rate. We used the floating chamber method to estimate the gas transfer velocity (k) of carbon dioxide (CO2), methane (CH4), and nitrous oxide (N2O) in the Markland Pool of the Ohio River, a large tributary of the Mississippi River (U.S.A). We measured k every two weeks for a year at one site and at 15 additional sites distributed across the length of the pool during two summer surveys. We found that k was positively related to both water currents and wind speeds, with 46% of the gas transfer attributable to water currents at low wind speeds (e.g., 0.5 m s−1) and 11% at higher wind speeds (e.g., >2.0 m s−1). Gas transfer velocity was highly sensitive to wind, possibly because the direction of river flow was often directly opposed to the wind direction. Gas transfer velocity values derived for CH4 were consistently greater than those derived for CO2 when standardized to a Schmidt number of 600 (k600), possibly because the transfer of CH4, a poorly soluble gas, was enhanced by surfacing microbubbles. Additional research to determine the conditions that support microbubble enhanced gas transfer is merited.