SUMMARY: Orange (Citrus sinensisL.) is one of the main fruit crops worldwide and its evergreen orchards may have a great potential for carbon (C) sequestration, but no data are currently available. In order to understand carbon fluxes in orange orchards, an experiment was undertaken on traditional and intensive planting systems.The experiment usedC. sinensisscions grafted ontoCitrus aurantium(bitter orange) rootstock. One orchard contained 14-year-old trees of the cv. Tarocco Scirè (a blood orange) grown in a traditional system with 494 trees/ha. The second orchard contained 12-year-old trees of the cv. Newhall (a seedless navel orange) grown in an intensive system with 1000 trees/ha. Net primary productivity (NPP) was obtained by measuring the annual canopy growth of single orange trees and the above ground dry biomass of the ground cover; soil respiration seasonal pattern was measured with an infrared gas analyser (EGM-4, PP System) from June 2005 to May 2006, every 2 weeks from 12·00 noon to 15·00 h for maximum respiration and from 02·00 to 05·00 h for minimum respiration; a 24 h cycle measurement of soil respiration was made every 3 months.Carbon fixation in the fruits and in the canopy of single trees was almost twice as much (10·7 kg C/tree) in the traditional than in the intensive system (5·5 kg C/tree); however, total NPP of the orchard did not change with planting density, being 5·3 t C/ha/year in the traditional system and 5·5 t C/ha/year in the intensive one. Carbon fixation by the ground cover was higher in the traditional (1·1 t C/ha/year) than in the intensive system (0·5 t C/ha/year). Annual soil respiration was 5·9 t C/ha/year in the traditional system and 4·2 t C/ha/year in the intensive one. The carbon balance was almost four times higher in the intensive system (1·8 t C/ha/year) than in the traditional one (0·5 t C/ha/year), due to large differences in soil respiration.