There has been much controversy, in the surgical treatment of idiopathic thoracolumbar and lumbar scoliosis, about whether anterior or posterior instrumentation produced a better result. This study compared the Zielke ventral derotation system (VDS) and Cotrel-Dubousset instrumentation (CDI) in the correction of frontal, sagittal, and roattional deformity, and defined the advantage and disadvantage of each instrument in idiopathic thoracolumbar and lumbar scoliosis. The ventral derotation system was used in 20 patients (VDS group) and Cotrel-Dubousset instrumentation was used in 20 patients (CDI group). The average age of VDS group patients at the time of surgery was 16.7 years, and that of CDI group patients was 18.5 years. The average length of follow up was 48 and 25 months, respectively. In frontal curve correction, the VDS group showed a 77% correction, and the CDI group a 63% correction (P > 0.05). In rotational deformity correction, the VDS group showed a 42% correction, and the CDI group a 16% correction (P<0.05). There were six decompensations in the VDS group, and three in the CDI group. Two decompensations of the VDS group were corrected with posterior CDI. There was one pseudarthrosis in the VDS group, with no significant symptoms. There were two instrumentation problems in each group, with no adverse effects on curve correction and maintenance. In conclusion, the ventral derotation system showed more powerful frontal and rotational correction, and can be effective even in more rigid or larger curves. Cotrel-Dubousset instrumentation, however, maintained physiologic sagittal alignment, was performed using a less invasive posterior approach, and had broader indication because it could be extended upward or downward to the compensatory curves.