STUDY DESIGN Fresh-frozen human cadaveric biomechanical study. OBJECTIVE To evaluate the biomechanical consequence of pedicle screw reinsertion in the thoracic spine. SUMMARY OF BACKGROUND DATA During pedicle screw instrumentation, abnormal appearance on fluoroscopic imaging or low current reading with intraoperatively evoked electromyographic stimulation of a pedicle screw warrants complete removal to reassess for pedicle wall violation or screw malposition. However, screw fixation strength has never been evaluated biomechanically after reinsertion using a previous pilot hole and trajectory. METHODS Thirty-one thoracic individual fresh-frozen human cadaveric vertebral levels were instrumented bilaterally with 5.5-mm titanium polyaxial pedicle screws, and insertional torque (IT) was measured with each revolution. A paired comparison was performed for each level. Screw reinsertion was performed by completely removing the pedicle screw, palpating the tract, and then reinserting along the same trajectory. Screws were tensile loaded to failure "in-line" with the screw axis. RESULTS There was no significant difference for pedicle screw pullout strength (POS) between reinserted and control screws (732 ± 307 N vs. 742 ± 320 N, respectively; P = 0.78). There was no significant difference in IT between initial insertion for the test group (INI) (0.82 ± 0.40 N·m) and control (0.87 ± 0.50 N·m) (P = 0.33). IT for reinserted screws (0.58 ± 0.47 N·m) had significantly decreased compared with INI and control screws (29% decrease, P = 0.00; 33% decrease, P = 0.00, respectively). The test group screws in the thoracic spine had significant correlations between initial IT and POS (r = 0.79, P = 0.00), and moderate correlations between reinsertion IT and POS in the thoracic spine (r = 0.56, P = 0.00). CONCLUSION Despite a significant reduction in pedicle screw IT, there was no significant difference in pedicle screw POS with reinsertion. Therefore, when surgeons must completely remove a pedicle screw for tract inspection, reinsertion along the same trajectory may be performed without significantly compromising fixation strength. LEVEL OF EVIDENCE N/A.