An experimental characterization of biogenic emission from Quercus ilex ssp. rotundifolia in a forest near Madrid, Spain, was carried out in the early autumn of the years 2000--2003. A dynamic branch enclosure technique was implemented to determine the monoterpene emission rates of this evergreen oak species during the 2000 and 2001 campaigns. Major compounds emitted during both measurement periods were limonene, α-pinene, ¿-pinene, sabinene, and myrcene. In the 2000 field campaign the light- and temperature-dependent model of Guenther et al. <1993> did not fit the data due to drastic reductions of emission rates (and leaf gas exchange related parameters) observed at high air temperature and low air humidity (high water vapor pressure deficit). This plant physiological activity depletion and the subsequent emission reduction were attributed to severe water soil deficit conditions, as precipitation was very scarce during the growing season. In contrast, during the 2001 field campaign, neither emission nor physiological activity showed strong decreases in hot days. A good fit of experimental data to Guenther model was achieved in this field campaign (r2 = 0.90), and linear regression gave a standard emission factor (ES) of 14.0 ¿g gdw-1 h-1 (gdw is grams dry weight). Soil moisture was presumably higher than during the 2000 campaign due to recent rain events. With the purpose of documenting the drought stress effect at canopy level, monoterpene oak fluxes were measured by the modified Bowen ratio micrometeorological technique throughout the 2001 field campaign and in the late summer of 2002 and 2003. The measured emission by both techniques showed a reasonably good correlation, although micrometeorological fluxes were, in general, lower than upscaled branch emission rates. According to Guenther's parameterization, standard emission fluxes (FS) of 0.30 ¿g m-2 s-1 (r2 = 0.61) and 0.28 ¿g m-2 s-1 (r2 = 0.67) were derived for the 2001 and 2002 field campaigns, respectively. However, measured fluxes in September 2003 at high ambient temperature were much lower than those of the previous years. Moreover, from the CLCT range in which emission in this field campaign could be fitted to a linear behavior, the FS value obtained was 50% lower than the 2001 and 2002 values. Water xylem potential measurements indicated the existence of drought conditions during the 2003 campaign and unstressed water conditions during the 2002 campaign.