Plantations play an important role in carbon sequestration and the global carbon cycle. However, there is a dilemma in that most plantations are managed on short rotations, and the carbon sequestration capacities of these short-rotation plantations remain understudied. Eucalyptus has been widely planted in the tropics and subtropics due to its rapid growth, high adaptability, and large economic return. Eucalyptus plantations are primarily planted in successive rotations with a short rotation length of 6~8 years. In order to estimate the carbon-stock potential of eucalyptus plantations over successive rotations, we chose a first rotation (FR) and a second rotation (SR) stand and monitored the carbon stock dynamics over a full rotation from 1998 to 2005. Our results showed that carbon stock in eucalyptus trees (TC) did not significantly differ between rotations, while understory vegetation (UC) and soil organic matter (SOC) stored less carbon in the SR (1.01 vs. 2.76 Mg.ha(-1) and 70.68 vs. 81.08 Mg. ha(-1), respectively) and forest floor carbon (FFC) conversely stored more (2.80 vs. 2.34 Mg. ha(-1)). The lower UC and SOC stocks in the SR stand resulted in 1.13 times lower overall ecosystem carbon stock. Mineral soils and overstory trees were the two dominant carbon pools in eucalyptus plantations, accounting for 73.77%~75.06% and 20.50%~22.39%, respectively, of the ecosystem carbon pool. However, the relative contribution (to the ecosystem pool) of FFC stocks increased 1.38 times and that of UC decreased 2.30 times in the SR versus FR stand. These carbon pool changes over successive rotations were attributed to intensive successive rotation regimes of eucalyptus plantations. Our eight year study suggests that for the sustainable development of short-rotation plantations, a sound silvicultural strategy is required to achieve the best combination of high wood yield and carbon stock potential.