This study is the first to directly correlate gas transfer velocity, measured at sea using the eddy-correlation (EC) technique, and satellite altimeter backscattering. During eight research cruises in different parts of the world, gas transfer velocity of dimethyl sulfide (DMS) was measured. The sample times and locations were compared with overpass times and locations of remote sensing satellites carrying Ku-band altimeters: ERS-1, ERS-2, TOPEX, POSEIDON, GEOSAT Follow-On, JASON-1, JASON-2 and ENVISAT. The result was 179 pairs of gas transfer velocity measurements and backscattering coefficients. An inter-calibration of the different altimeters significantly reduced data scatter. The inter-calibrated data was best fitted to a quadratic relation between the inverse of the backscattering coefficients and the gas transfer velocity measurements. A gas transfer parameterization based on backscattering, corresponding with sea surface roughness, might be expected to perform better than wind speed-based parameterizations. Our results, however, did not show improvement compared to direct correlation of shipboard wind speeds. The relationship of gas transfer velocity to satellite-derived backscatter, or wind speed, is useful to provide retrieval algorithms. Gas transfer velocity (cm/hr), corrected to a Schmidt number of 660, is proportional to wind speed (m/s). The measured gas transfer velocity is controlled by both the individual water-side and air-side gas transfer velocities. We calculated the latter using a numerical scheme, to derive water-side gas transfer velocity. DMS is sufficiently soluble to neglect bubble-mediated gas transfer, thus, the DMS transfer velocities could be applied to estimate water-side gas transfer velocities through the unbroken surface of any other gas.