Objective: To evaluate the cases of osteoporotic compression fracture with kyphotic deformity treated with modified pedicle subtraction osteotomy and analyze the usefulness of this osteotomy.Methods: Twenty patients (nine men, 11 women; mean age, 66.4 years; mean follow-up duration, 39.6 months) who underwent modified pedicle subtraction osteotomy at our hospital between November 2003 and July 2012 with a minimum follow-up period of 2 years were included. All cases were injuries induced by a fall incurred while standing or lifting heavy objects without high-energy trauma. The mean bone mineral density-dual energy x-ray absorptiometry (BMD-DEXA) was 0.673 g/cm 2 (0.571-0.740 g/cm 2 ), which was -2SD below normal, indicating severe osteoporosis. The indications for surgery included gait disturbance due to severe pain with pseudarthrosis, increased kyphotic angle, and progressive neurological symptoms. We evaluated Japanese Orthopaedic Association scores, Oswestry Disability Index scores, bone union, change in kyphotic angle, estimated blood loss, complications, and the patient's overall clinical satisfaction.Results: The collapsed vertebrae were thoracic (T 9, 11, 12 ) in 13 cases and lumbar (L 1 ) in seven cases. The mean pre-and post-operative kyphotic angles were 26.2 AE 9.9 degrees and 8.3 AE 8.1 degrees, showing an 18.3-degree correction. Plain radiography at the last follow-up showed a mean angle of 17.9 AE 12.6 degrees, indicating a 9.5-degree correction loss. The overall correction was 8.2 degrees. Bony union was achieved at the last follow-up in all cases. Clinical outcomes showed improvement in mean Japanese Orthopaedic Association score from 13.2 to 21.7 and mean Oswestry Disability Index score from 40.3 to 13.6. Overall clinical patient satisfaction showed 12 excellent results, five moderate results, and three poor results. Mean operation time was 4.2 h (range, 4-6 h). The overall mean estimated blood loss (EBL) was 1098 mL (range, 750-1370 mL). The mean hospital stay was 3.6 weeks (range, 2-6 weeks). There was one case of cauda equina syndrome at 2 days postoperatively, and two cases of screw loosening. There were no cases of distal junctional kyphosis, but there were two cases of proximal junctional kyphosis.Conclusions: Despite limited correction of kyphotic angle, our modified pedicle subtraction osteotomy technique resulted in satisfactory outcomes in our patients, who had persistent severe back pain, increased kyphotic angle with pseudarthrosis, and progressive neurologic deficits. Nevertheless, this procedure requires careful follow-up and strict surgical indications because of the risk of neurological damage and technical problems.