An intensive investigation of particulate sulfur fluxes to a deciduous forest was conducted at Oak Ridge, Tennessee, during May 1983, as part of a multilaboratory program to study the dry deposition of airborne trace gases and aerosols to vegetated surfaces. At this time, the leaf area was approximately 50% of the maximum and was rapidly increasing. Eddy correlation methods were used to measure fluxes of submicron particulate sulfur (presumably sulfate), of submicron particles in three different size ranges, and of ozone above the forest canopy. The measured deposition velocities (vd) of particulate sulfur peaked at about 1 cm s-1 in daytime, but dropped to near zero at night, with a long-term average of about 0.6 cm s-1. When scaled with friction velocity, these values of vd are consistent with similar measurements taken previously over different surfaces. Measurements and analyses indicate that the flux divergence of particulate sulfur should have caused an error no greater than ¿0.05 cm s-1 in vd estimates. For very small particles (less than 0.1 μm diameter), measured values of vd were similar to those for particulate sulfate, but deposition velocities for particles in the accumulation size range decreased as particle size increased, and upward components of fluxes were found in daytime convective conditions. During this study, measurements were also made of sulfur deposition to polycarbonate petrie dishes placed in the forest canopy, and to the bucket of a standard wet-dry collector in a forest clearing. These surrogate surface studies produced much smaller values of vd for sulfate than those derived from eddy correlation, when no scaling as a function of leaf area relative to ground area was applied. ¿ American Geophysical Union 1989