<h3>BACKGROUND CONTEXT</h3> Posterior cervical fusion (PCF) is successfully used to treat cervical spondylotic myelopathy, multilevel cervical stenosis, and cervical deformity. However, limited evidence exists regarding the appropriate level of proximal and distal extension of PCF constructs. Further investigation is required to understand the effects of construct endpoint on fusion rate, junctional failure, and sagittal deformity correction. <h3>PURPOSE</h3> To compare the surgical and radiographic outcomes of patients undergoing posterior cervical fusion (PCF) with constructs extending from C2-T2 to patients with constructs extending from C3-T1. <h3>STUDY DESIGN/SETTING</h3> This was a multicenter retrospective review of two prospective cohorts of patients undergoing PCF from C2-T2 or C3-T1 for subaxial cervical stenosis at two academic institutions from 2012 to 2020. <h3>PATIENT SAMPLE</h3> We identified 106 patients for inclusion in the C2-T2 cohort and 49 patients for inclusion in the C3-T1 cohort. <h3>OUTCOME MEASURES</h3> Pre- and postoperative cervical alignment parameters, distal screw loosening, distal junctional failure, and pseudarthrosis were measured. <h3>Methods</h3> Cervical alignment parameters were measured on both preoperative and postoperative radiographs performed greater than 6 months postoperatively. The cohorts were compared based on preoperative cervical alignment, postoperative cervical alignment, and change in cervical alignment. Postoperative radiographs were also assessed for signs of distal screw loosening, distal junctional failure, and pseudarthrosis. We utilized Student's t-test to compare all means between groups with p<0.05 deemed statistically significant. <h3>Results</h3> A total of 155 patients were included in the study (C2-T2: 106 patients, C3-T1: 49 patients). There were no significant differences in demographics or preoperative symptoms between cohorts. Fusion rates were significantly higher in the C2-T2 (93%) than the C3-T1 (80%, p=0.040) cohort. When comparing the C2-T2 to the C3-T1 cohort, the C3-T1 cohort had a significantly greater rate of proximal junctional failure (2% vs 10%, p=0.006), distal junctional failure (1% vs 20%, p<0.001) and distal screw loosening (4% vs 15%, p=0.02). Although ΔC2-7 Sagittal vertical axis increased significantly in both cohorts (C2-T2: 6.2°, p=0.04; C3-T1: 8.4°, p<0.001), correction did not significantly differ between groups (p=0.32). The C3-T1 cohort had a significantly greater increase in ΔC2 Slope (8.0° vs 3.1°, p=0.03) and ΔC0-C2 Cobb angle (6.4° vs 1.2°, p=0.04) <h3>Conclusions</h3> In patients undergoing PCF, a C2-T2 construct demonstrated lower rates of pseudarthrosis, DJF, PJF, and compensatory upper cervical hyperextension compared to a C3-T1 construct. <h3>FDA DEVICE/DRUG STATUS</h3> This abstract does not discuss or include any applicable devices or drugs.