The air-sea gas transfer velocity k is frequently estimated as an empirical function of wind speed. However, it is widely recognized that k depends on processes other than wind speed alone. The small-eddy model, which describes periodic events of small eddies disturbing the sea surface with water from below, suggests a direct relation between k and the dissipation rate of turbulent kinetic energy ϵ at the air-sea interface. This relation has been proven both in laboratories and in the field in various freshwater and coastal environments, but to date has not been verified in open ocean conditions. Here, concurrent North Atlantic field observations of ϵ and eddy covariance measurements of DMS and CO2 air-sea gas flux are presented. Using ϵ measurements, we compare the small-eddy model at various depths to previously published observations. Extrapolating the measured ϵ profiles to the thickness of the viscous sublayer allows us to formulate a function of k that depends solely on the water side friction velocity u*w, which can be inferred from direct eddy covariance measurements of the air-side friction velocity u*a. These field observations are generally consistent with the theoretical small-eddy model. Utilizing a variable Schmidt number exponent in the model, rather than a constant value of 1/2 yields improved agreement between model and observations.