Purpose : The purpose of this study was to compare trunk repositioning errors between subjects with and without low back pain in sitting and standing. Methods : Total 81 participants were recruited who consisted of 41 subjects with low back pain and 40 normal subjects. The subjects were instructed to replicate the predetermined target positions of the trunk toward upright and $30^{circ}$ flexion in sitting and standing. During each of movement, digital inclinometer was used to measure the angular movement of $T_{12}$ spinal process. Repositioning error was calculated as the absolute difference between the predetermined target positions and replicated target positions. Results : In subjects with low back pain, upright repositioning error was $1.26^{circ}{pm}0.14^{circ}$ in sitting and $1.55^{circ}{pm}0.24^{circ}$ in standing, and $30^{circ}$ flexion repositioning error was $3.23^{circ}{pm}0.33^{circ}$ in sitting and $5.50^{circ}{pm}0.50^{circ}$ in standing. In subjects without low back pain, upright repositioning error was $1.38^{circ}{pm}0.15^{circ}$ in sitting and $1.67^{circ}{pm}0.18^{circ}$ in standing, and flexion repositioning error was $2.61^{circ}{pm}0.28^{circ}$ in sitting and $3.70^{circ}{pm}0.52^{circ}$ in standing. It was demonstrated that flexion repositioning error increased significantly in standing position. In subjects with low back pain, $30^{circ}$ flexion repositioning error was significantly higher in standing than in sitting. Conclusion : The repositioning error of subjects with low back pain increased during flexion and it implies that some aspects of proprioception are decreased in subjects with low back pain. Therefore, it will be emphasis that a clinical trial to increase the trunk flexion stability of subjects with low back pain in standing.