Three approaches commonly used to quantify diffusive gas exchange across aquatic surfaces were compared in a densely treed, low-wind environment Diffusive surface fluxes of carbon dioxide (CO2) and methane (CH4) from a small boreal reservoir were estimated using (i) surface water concentrations, the thin boundary layer (TBL) equation, and gas transfer velocities (k) calculated using sulfur hexafluoride (SF6); (ii) surface water concentrations, the TBL equation, and k estimated from wind speed; and (iii) static floating chambers (FCs). Comparisons were made during three different approximately 10-day intervals (August 2000, June and September 2001). CO2 and CH4 fluxes estimated from SF6-derived k were on average 1-3 times greater than those determined from wind-estimated k Overall agreement between FC CO2 and CH4 flux estimates and those based on SF6 and wind speed derived kvalues was much weaker, with FC CO2 and CH4 flux estimates ranging from -9 to 23 times those based on SF6 and wind-estimated k values. Chamber deployment likely enhanced gas transfer through disturbance of the surface boundary layer, and results of this study suggest that caution must be exercised concerning the use of FCs on very still water surfaces. Furthermore, findings of this study contradict the common belief that use of wind speed to approximate k is inappropriate for small bodies of water characterized by low winds and surface obstructions.