Nitrogen (N) addition enhances primary productivity of terrestrial ecosystems. However, the effects of N fertilization and/or deposition on net ecosystem CO2 exchange (NEE) are not fully understood. The effects of N on NEE were investigated in two experimental cheatgrass ecosystems in Ecologically Controlled Enclosed Lysimeter Laboratories (EcoCELLs), Reno, Nevada. In this experiment, no N fertilization was added to the two EcoCELLs in the first year and two different N fertilization regimes were applied in the second year. N fertilizer was applied once to one EcoCELL (pulse fertilization, PF). and the same total amount of N in biweekly increments to the other EcoCell (gradual fertilization, GF). NEE, photosynthetically active radiation (PAR) and canopy green leaf area index (LAI) were continuously measured in the two EcoCELLs during the pretreatment and N-fertilized years. Plant N content and biomass were measured at the end of the growing season in each year. Radiation-use efficiency (RUFCO2) was calculated as the ratio of gross ecosystem photosynthesis (GEP) to the intercepted photosynthetically active radiation (IPAR). The responses of NEE to IPAR were used to estimate the maximum ecosystem photosynthetic capacity (F-max). N fertilization stimulated canopy LAI, plant N content, F-max, RUECO2, NEE and biomass in both methods of N supply applications. PF led to higher LAI, F-max and NEE than GF, but both had a similar RUECO2 during the early growing season. GF maintained higher LAI, F-max, RUECO2 and NEE than PF during the late growing season. At the ecosystem level, N fertilization stimulated daily NEE directly by increasing canopy LAI, plant N content, shoot/root ratio and the maximum ecosystem photosynthetic capacity, and increased the seasonally accumulated NEE indirectly by extending the growing season. PF differed significantly from GF in its effects on NEE and RUECO2, possibly due to differential rates and timing of N availability. Our study suggested that these changes in the canopy RUECO2 and growing season under N fertilization or N deposition regimes should be considered in modeling studies of ecosystem C sequestration. (C) 2009 Elsevier B.V. All rights reserved.