CONTEXT: Open-door laminoplasty often results in postoperative complications such as loss of cervical lordosis, limitations of cervical motion, and axial symptoms. However, current modified laminoplasty techniques such as muscle-sparing type or spinous process splitting technique are not as effective as expected. PURPOSE: To evaluate the radiological and clinical outcomes of C3 laminectomy combined with modified unilateral laminoplasty (preservation of posterior muscle-ligament complex and reconstruction of the midline structures) versus traditional open door laminoplasty in treating cervical spondylotic myelopathy or ossification of the posterior longitudinal ligament. STUDY DESIGN: Retrospective case-control study. PATIENT SAMPLE: One hundred and eleven patients with multilevel cord compression and myelopathic symptoms. OUTCOME MEASURES: The outcome parameters were operation time, blood loss volume, complications, osseous fusion status, C0-C2 and C2-C7 Cobb angles, T1 slope, cervical sagittal vertical axis (cSVA), cervical curvature index (CCI), range of motion (ROM), cross-sectional area (CSA) of the semispinalis cervicis, axial symptoms, visual analog scale (VAS) score, Japanese Orthopedic Association (JOA) score, and neck disability index (NDI). METHODS: We matched 37 patients who underwent modified laminoplasty with 74 patients treated by traditional open door laminoplasty (ratio, 1:2) according to age, sex, body mass index, compromised level, and radiographic characteristics. Preoperative and postoperative cervical parameters, namely, the C2-C7 Cobb angle, ROM, and CCI were measured on X-ray films. The CSA of the semispinalis cervicis was assessed on magnetic resonance images, and osseous fusion status of the hinge side and the osteotomy site was evaluated by computed tomography. We used the JOA and VAS scores, and the NDI to evaluate clinical outcomes. RESULTS: The average follow-up period in the modified group was 24.1 months (range, 18-37 months) compared with 24.7 months (range, 18-38 months) in the control group. At the final follow-up, C0-C2 Cobb angle, T1 slope, and cSVA increased in the control group and were unchanged in the modified group. The C2-C7 Cobb angle decreased significantly in the control group and did not change in the modified group. ROM and CCI loss rate did not change in the modified group but decreased significantly in the control group. The CSA loss in the semispinalis cervicis was 222.90 +/- 79.56 mm2 in the control group and 49.11 +/- 75.93 mm(2) in the modified group, with a significant difference (p<.001). The final CSA of the semispinalis cervicis at C2 and C4-C7 levels showed no significant difference in the modified group and decreased significantly in the control group compared with preoperation. Changes in the C2-C7 Cobb angle and cSVA were both correlated with the CSA loss of the semispinalis cervicis (r=0.282, p=.003; r=0.267, p=.005, respectively). Moreover, the CSA loss of the semispinalis cervicis also conelated with the CCI loss rate and the changes in ROM (r=0.312, p=.001; r=0.287, p=.002, respectively). Clinical outcomes such as VAS and NDI scores, improved significantly more in the modified group versus the controls (p<.001 and p=.005, respectively), while JOA scores improved similarly in both groups (p=.132). The incidence of axial symptoms was significantly lower in the modified group versus controls (5.4% vs 9.5%, respectively; p=.023). CONCLUSIONS: C3 laminectomy combined with modified unilateral laminoplasty is effective for treating patients with multilevel cord compression. This modified technique reconstructs the midline structures and may lead to improved alignment and less axial pain. (C) 2020 Published by Elsevier Inc.