The dynamic-chamber technique was used to investigate the correlation between NH $ and NO # fluxes and different climatic and physiological parameters : air temperature ; relative air humidity ; photosynthetic photon fluence rate ; NH $ and NO # concentrations ; transpiration rate ; leaf conductance for water vapour ; and photosynthetic activity. The experiments were performed with twigs from the sun crown of mature beech trees (Fagus sylvatica) at a field site (Ho$ glwald, Germany), and with 12-wk-old beech seedlings under controlled conditions. Both sets of experiments showed that NO # and NH $ fluxes depended linearly on NO # and NH $ concentration, respectively, in the concentration ranges representative for the field site studied, and on watervapour conductance as a measure for stomatal aperture. The NO # compensation point determined in the field studies (the atmospheric NO # concentration with no net NO # flux) was 1n8-1n9 nmol mol −" . The NH $ compensation point varied between 3n3 and 3n5 nmol mol −" in the field experiments, and was 3n0 nmol mol −" in the experiments under controlled conditions. The climatic factors T and PPFR were found to influence both NO # and NH $ fluxes indirectly, by changing stomatal conductance. Whilst NO # flux showed a response to changing relative humidity that could be explained by altered stomatal conductance, increased NH $ flux with increasing relative humidity ( 50 %) depended on other factors. The exchange of NO # between above-ground parts of beech trees and the atmosphere could be explained exclusively by uptake or emission of NO # through the stomata, as indicated by the quotient between measured and predicted NO # conductance of approx. 1 under all environmental conditions examined. Neither internal mesophyll resistances nor additional sinks could be observed for adult trees or for beech seedlings. By contrast, the patterns of NH $ flux could not be explained by an exclusive exchange of NH $ through the stomata. Deposition into additional sinks on the leaf surface, as indicated by an increase in the quotient between measured and predicted NH $ conductance, gained importance in high air humidity, when the stomata were closed or nearly closed and\or when atmospheric NH $ concentrations were high. Although patterns of NH $ gas exchange did not differ between different months or years at high NH $ concentrations (c. 140 nmol mol −" ), it must be assumed that emission or deposition fluxes at low ambient NH $ concentration (0n8 and 4n5 nmol mol −" ) might vary significantly with time because of variation in the NH $ compensation point.