To upgrade existing activated sludge treatment plants, different techniques that would remove an important flux of nitrogen rapidly on a great number of units were investigated. Nitrification with conventional activated sludge systems requires considerable multiplication of tankage volume. The necessary investment and space is not always available, especially since many older plants are now in urbanized areas. To lower the nitrogen load in receiving water, the first priority should be to obtain partial nitrogen removal with existing plants, using methods that are simple to adapt.Three techniques were tested on large scale: submerged elements in aeration basin to add fixed biomass, contact stabilisation that allows a higher sludge age in the same tankage volume, and adding submerged biotower packings as a tertiary aeration stage. In a full scale unit (4000 m3/d), one complete section of the plant fitted with biofilter packing was operated in parallel with a similar unmodified section as reference. The volume occupied by the fixed beds was varied between 20 and 40 % of the tank. The submerged elements improved removal efficiency, to maintain effluent quality at higher loadings or obtain lower residual pollution values in existing plants. The biofilm evolution and the hydraulic behaviour of the packing was followed. No significant change in sludge settleability was observed, but fixed biomass addition reduced sludge production because of a lower overall mass loading. The resulting higher sludge age allowed the ammonia oxidizers to remain in the mixed population beyond usual F/M limits, but no installation of nitrifiers on the support media could be observed. To verify the limits of immersed plastic surfaces for nitrification, an aerated column was fed with effluent of a highly loaded activated sludge plant. In opposition to carriers submerged in mixed liquor, nitrifier attachment was obtained, and COD and SS removal for effluent polishing was achieved. With a carbon loading exceeding 1,5 kg COD/m3 d, a maximum oxidation rate of 0,4 kg N-NH4/m3 d could be obtained. A pilot unit was tested to assess the potential volume reduction for nitrogen elimination by contact stabilisation. This configuration stores the highly concentrated return sludge in a reaeration basin, and keeping only the minimum detention time in the contact basin to obtain nitrification. Also, an increased carbon load in the contact basin enhances denitrification. For urban wastewaters with a COD/N ratio of about 10, complete oxidation and partial removal of nitrogen were obtained with a volume loading of 1,5 kg COD/m3 d. Nitrogen removal rates of 0,15 kg N/m3 d were measured both in the anoxic and the aerobic part of the contact basin. The contact stabilisation mode was then tested on full scale combined with submerged biomass carriers. A consistant nitrogen elimination of 50 % was obtained with aeration detention times of about 4 hours.