The long-term responses of trees to elevated CO2 are especially crucial (1) to mitigating the rate of atmospheric CO2 increase, (2) to determining the character of future forested natural ecosystems and their spread across the landscape, and (3) to determining the productivity of future agricultural tree crops. Therefore, a long-term CO2-enrichment experiment on sour orange trees was started in 1987, and the final results after 17 years are reported herein. Four sour orange trees (Citrus aurantium L.) were grown from seedling stage at 300 μmol mol−1 CO2 above ambient in open-top, clear-plastic-wall chambers at Phoenix, AZ. Four control trees were similarly grown at ambient CO2. All trees were supplied ample water and nutrients comparable with a commercial orchard. After a peak 2–4 years into the experiment, there was a productivity plateau at about a 70% enhancement of annual fruit and incremental wood production over the last several years of the experiment. When summed over the duration of the experiment, there was an overall enhancement of 70% of total biomass production. Much of the enhancement came from greater numbers of fruits produced, with no change in fruit size. Thicker trunks and branches and more branches and roots were produced, but the root/shoot ratio was unaffected. Also, there was almost no change in the elemental composition of the biomass produced, perhaps in part due to the minimal responsiveness of root-symbiotic arbuscular mycorrhizal fungi to the treatment.