Aqueous evaporation and condensation kinetics are poorly understood, and uncertainties in their rates affect predictions of cloud behavior and therefore climate. We measured the cooling rate of 3 M ammonium sulfate droplets undergoing free evaporation via Raman thermometry. Analysis of the measurements yields a value of 0.58 ؎ 0.05 for the evaporation coefficient, identical to that previously determined for pure water. These results imply that subsaturated aqueous ammonium sulfate, which is the most abundant inorganic component of atmospheric aerosol, does not affect the vapor-liquid exchange mechanism for cloud droplets, despite reducing the saturation vapor pressure of water significantly.aerosol ͉ clouds ͉ aqueous ͉ inorganic T he vapor-liquid exchange dynamics of water underlie vital processes in biology, engineering, and atmospheric science. The evaporation and condensation rates of water are particularly important in the formation of cloud particles and are among the largest unknowns in assessing the impact of indirect aerosol effects on the radiative balance in the atmosphere (1). Model studies suggest that evaporation rates Ͻ10% of the maximum rate determined by gas kinetic theory for atmospherically relevant aqueous systems would indicate kinetic control over cloud growth processes, with implications for cloud and aerosol models (2, 3). Attempts to quantify the evaporation and condensation rates of water have yielded values spanning three orders of magnitude, although the most recent values converge to a single order of magnitude (4-10). This uncertainty contributes to the wide variability in cloud model predictions for the anthropogenic effect on size and number of cloud condensation nuclei (CCN) and the corresponding radiative impact on the global system (1,3,11,12).Measuring evaporation and condensation rates of liquid water is difficult because of the complex heat and mass transfer processes occurring in such experiments. The temperature of the liquid surface must be accurately determined and, in most cases, both evaporation and condensation must be properly accounted for. In addition, it has been suggested that some of the earlier measurements of these rates, which typically involved static liquid surfaces, yielded artificially low rates because of the contamination of the liquid with surface impurities (6). Such impurities may well be present in atmospheric conditions; indeed, field measurements of aerosol growth rates appear to indicate such a slowing effect (13,14). In our previous work, we addressed the problem of accurately modeling the heat and mass transfer processes through our measurements of the free evaporation of pure H 2 O and pure D 2 O (4, 5, 10). We also studied relative evaporation rates of isotopomers in mixtures (10,15). In those studies of pure solutions, condensation was negligible, allowing the evaporation process to be modeled accurately and then directly related to the cooling rate of the droplets. We found that evaporation for both pure H 2 O and pure D 2 O occurred at Ϸ60%...