OBJECTIVE Thoracic pedicle screw fixation has many advantages over other methods of spinal fixation in treating various conditions. We evaluated the safety and efficacy of pedicle screw fixation in the thoracic spine. METHODS We retrospectively reviewed 144 patients (65 males, 79 females; mean age 38 years; range 9 to 82 years) who underwent posterior thoracic pedicle screw instrumentation. A total of 827 thoracic pedicle screws were inserted to the thoracic spine (T1-12) by the same senior spine surgeon using the free-hand technique. Indications for thoracic spine surgery were deformities (n=47, 32.6%), metastatic or primary tumors (n=15, 10.4%), spinal infections (n=7, 4.9%), degenerative diseases (n=13, 9%), and spinal trauma (n=62, 43.1%). Screw containment was assessed by three independent reviewers on postoperative plain radiographs. In addition, thin-slice computed tomography scans were obtained in 21 suspected cases (29 screws). RESULTS Screw containment was found in 94.3% (780 screws). Incorrect screw placement was found in 47 screws (5.7%), including 31 screws (3.8%) demonstrated by plain radiographs, and 16 screws (1.9%) demonstrated by thin-slice computed tomography scans. More than half of the faulty screws (n=24, 51.1%) were detected in scoliosis patients, especially on the concave side of the curvature (83.3%). The most frequent site of faulty screw placement was the T8 level (21.3%). No symptoms or complications occurred related to faulty screw placement. CONCLUSIONS The high accuracy of thoracic pedicle screw placement (94.3%) using the free-hand technique suggests that pedicle screw fixation of the thoracic spine is safe, reliable, and useful in the treatment of all types of spinal diseases.