New micrometeorological measurements of small (0.1--0.2 ¿m diameter) aerosol particle fluxes using the eddy correlation technique are presented for moorland and also for grassland vegetation, the latter measurements being made both before and after cutting of the grassland to observe the resultant change in particle deposition velocity. These data are considered together with previously reported and reanalyzed micrometeorological measurements, again using the eddy correlation technique, for a number of different surface types, including arable crops and forest. Differences in observed surface deposition velocities, vds, due to the different surface roughnesses are highlighted. It was found that the various data sets showed a wholly consistent behavior when ensemble averages over the typical range of atmospheric stability ranges are considered in order to reduce the scatter inherent in these types of measurements. A working parameterization of surface deposition velocity in terms of the surface's roughness length, zo, is presented. This is then extended for different atmospheric stabilities, using the parameterization suggested by Lamaud et al. <1994c>, to yield vds/u* = k1 + k2 (-300 z/L)2/3, where k1 = k1 = 0.001222 log(zo) + 0.003906, k2 = 0.0009, where z is the measurement height, L is the Obukhov stability length, and u* is the local friction speed. The new data are finally compared to current analytical model descriptions of the deposition process, highlighting deficiencies in our understanding of the surface collection efficiency even for these small particles.