Biomechanical modeling estimated that trunk muscle activity during various tasks could be reduced by 1-14% without the loss of spine stability when a lumbosacral orthosis (LSO) is worn [Cholewicki, J., 2004. The effects of lumbosacral orthoses on spine stability: what changes in EMG can be expected? Journal of Orthopedic Research 22, 1150-1155]. The present study experimentally tested these theoretical predictions in an unstable sitting task. This task required subjects to balance on a seat supported by a plastic hemisphere (slashed circle=30cm) and placed on a force plate that tracked the center of pressure (CoP). The average CoP velocity quantified subjects' performance. Healthy subjects (12 males, 11 females) balanced for 20s in 3 trials performed with and without the LSO in random order. EMG was recorded bilaterally from rectus abdominis (RA), external oblique (EO), thoracic (TES) and lumbar erector spinae (LES), and expressed as the % of maximum voluntary activation (%MVA). There was no difference in the balance performance with and without the LSO (p=0.13). However, EMG averaged across the trials was significantly lower in the LSO, as compared to the No LSO condition, for TES (5.8+/-3.2 vs. 6.4+/-3.7%MVA, p=0.02) and LES (3.7+/-1.5 vs. 5.9+/-3.9%MVA, p=0.01). No significant differences were present in the abdominal muscle activity. These results agree with earlier spine modeling simulations, which predicted the greatest reduction in muscle activity due to LSO to occur in TES and LES. It was hypothesized that such a reduction in muscle co-contraction could benefit patients with low back pain, who exhibit elevated muscular activity during postural tasks such as walking, standing and sitting.