Real-time total gaseous sulfur concentrations were measured (sampling frequency 1 Hz) from the R/V Oceanus as part of the Atlantic Stratocumulus Transition Experiment/Marine Aerosol and Gas Exchange (ASTEX/MAGE) field campaign in the Azores during June of 1992. The measurements were used to estimate sulfur gas sea-to-air flux, determine the size scale of sulfur gas inhomogeneities in the marine boundary layer, and the timescale of sampling necessary to characterize a meaningful air mass. Using the time scale of sampling and wind speed measurements, the size scale of sampling can also be determined. Sea-to-air sulfur gas flux estimates were obtained using a variance method and the inertial-dissipation method. Values from five 1-hour measurement periods on two different days ranged from 21 to 28 μmol/m2 d and 11 to 17 μmol/m2 d, respectively. These values are above the range of 1 to 13 μmol/m2 d reported by Blomquist et al. (this issue) during ASTEX/MAGE, based on the dimethyl sulfide (DMS) concentration in surface seawater, using the stagnant boundary layer model of air-sea exchange. A power spectrum of the sulfur gas time series shows the contribution of each turbulent eddy size to the total signal variance. In a representation of the power spectrum, the area under any portion of the curve is proportional to variance. From these power spectra we have obtained the fraction of the total variance in ground level sulfur concentration fluctuations as a function of time. The data indicate that air samples should be integrated for 1000 s for ground-based measurements. ¿ American Geophysical Union 1996