Long-term fluxes of ozone (O(3)) were measured over a mixed temperate forest using the aerodynamic gradient method. The long-term average O(3) flux (F) was -366 ng m(-2) s(-1) for the period 2000-2010, corresponding to an average O(3) concentration of 48 μg m(-3) and a deposition velocity v(d) of 9 mm s(-1). Average nocturnal ozone deposition amounted to -190 ng m(-2) s(-1), which was about one third of the daytime flux. Also during the winter period substantial O(3) deposition was measured. In addition, total O(3) fluxes were found to differ significantly among canopy wetness categories. During the day, highest deposition fluxes were generally measured for a dry canopy, whereas a rain-wetted canopy constituted the best sink at night. Flux partitioning calculations revealed that the stomatal flux (F(s)) contributed 20% to the total F but the F(s)/F fraction was subject to seasonal and diurnal changes. The annual concentration-based index AOT40 (accumulated dose over a threshold of 40 ppb) and the Phytotoxic Ozone Dose (POD(1) or accumulated stomatal flux above a threshold of 1 nmol m(-2) s(-1)) were related in a curvilinear way. The O(3) deposition was found to be largely controlled by non-stomatal sinks, whose strength was enhanced by high friction velocities (u(*)), optimizing the mechanical mixing of O(3) into the canopy and the trunk space. The long-term geometrical mean of the non-stomatal resistance (R(ns)) was 136 s m(-1) but lower R(ns) values were encountered during the winter half-year due to higher u(*). The R(ns) was also subject to a marked diurnal variability, with low R(ns) in the morning hours, when turbulence took off. We speculate that non-stomatal deposition was largely driven by scavenging of ozone by biogenic volatile organic compounds (BVOCs) and especially NO emitted from the crown or the forest floor.