Atmosphere-surface exchange represents one mechanism by which atmospheric particle mass and number size distributions are modified. Deposition velocities (upsilon(d)) exhibit a pronounced dependence on surface type, due in part to turbulence structure (as manifest in friction velocity), with minima of approximately 0.01 and 0.2 cm s(-1) over grasslands and 0.1-1 cm s(-1) over forests. However, as noted over 20 yr ago, observations over forests generally do not support the pronounced minimum of deposition velocity (upsilon(d)) for particle diameters of 0.1-2 mu m as manifest in theoretical predictions. Closer agreement between models and observations is found over less-rough surfaces though those data also imply substantially higher surface collection efficiencies than were originally proposed and are manifest in current models. We review theorized dependencies for particle fluxes, describe and critique model approaches and innovations in experimental approaches, and synthesize common conclusions of experimental and modelling studies. We end by proposing a number of research avenues that should be pursued in to facilitate further insights and development of improved numerical models of atmospheric particles.