The static closed chamber technique is used in the study on the CH4 and N2O fluxes from the soils of primeval Abies fabri forest, the succession Abies fabri forest and the clear-cut areas of mid-aged Abies fabri forest in the Gongga Mountain from May 1998 to September 1999. The results indicate the following: (i) The forest soil serves as the source of atmospheric N2O at the three measurement sites, while the fluxes of CH4 are all negative, and soil is the sink of atmospheric CH4. The comparative relations of N2O emissions between the three sites are expressed as primeval Abies fabri forest > clear-cut areas > succession Abies fabri forest, and those of CH4 consumption fluxes are primeval Abies fabri forest > succession Abies fabri forest > clear-cut areas. (ii) Signifi-cant seasonal variations of N2O emission at various sites were observed, and two emission peaks of N2O occurr during summer (July—August) and spring (February—March), whereas N2O emission is relatively low in winter and spring (mid March—April). Seasonal variations of CH4 consumption at each measurement site fluctuate drastically with unclear regularities. Generally, CH4 consumption fluxes of succession Abies fabri forest and clear-cut areas are higher from mid May to late July but lower in the rest of sampling time, while the CH4 flux keeps a relatively high value even up to Sep-tember in primeval Abies fabri forest. In contrast to primeval Abies fabri forest, the CH4 absorbabili-ties of succession Abies fabri forest and clear-cut areas of mid-aged Abies fabri forest are weaker. Particularly, the absorbability of the clear-cut areas is even weaker as compared with the other two sites, for the deforestation reduces the soil absorbability of atmospheric CH4. (iii) Evident diurnal variation regularity exists in the N2O emissions of primeval Abies fabri forest, and there is a statistic positive correlation between the fluxes of N2O and air temperature (R=0.95, n=11, <0.01), and also the soil temperature of 5-cm layer (R=0.81, n=11, < 0.01), whe
