Objective The study aimed to investigate the correlation between the severity of disease and postoperative neurological recovery in patients with cervical spondylotic myelopathy (CSM) combined with developmental spinal stenosis. Methods A retrospective analysis of the clinical data of 114 CSM patients combined with developmental spinal stenosis admitted to our hospital from June 2019 to June 2020 was performed. All of the patients who underwent posterior cervical unidoor vertebroplasty were divided into the mild, moderate, and severe groups according to the Torg–Pavlov ratio. The clinical data including patients' age, course of spinal cord high signal change, and first onset age were collected. The recovery time, preoperative, and postoperative Japanese Orthopaedic Association (JOA) scores of patients in each group were compared with the calculation of the improvement rate. The correlation between the severity of disease and postoperative neurological recovery in CSM patients combined with developmental spinal stenosis was analyzed by Pearson correlation. The factors influencing postoperative neurological recovery were analyzed using multivariate logistic regression analysis. The receiver operating characteristic curve (AUC) was used to evaluate the value of each influencing factor in predicting postoperative recovery. Results Significant differences were observed in the proportion of linear hyperintensity changes in the spinal cord, the age of first onset, the course of the disease, and the Torg–Pavlov ratio among the mild, moderate, and severe groups (P < 0.05). The postoperative recovery time of the moderate and severe groups was significantly higher than that of the mild group, while the preoperative JOA score was significantly lower than that of the mild group. On the other hand, the postoperative recovery time of the severe group was prominently higher than that of the moderate group, whereas the preoperative JOA score was observably lower than that of the moderate group (P < 0.05). Pearson correlation analysis showed that the postoperative recovery time was significantly negatively correlated with the Torg–Pavlov ratio, age at first onset, and disease course (r = −0.359, −0.502, −0.368, P < 0.05), while it was positively correlated with spinal cord linear high-signal changes (r = 0.641, P < 0.05). Multifactorial logistic regression analysis revealed that the Torg–Pavlov ratio, age at first onset, and disease course were protective factors, while spinal cord linear high-signal alterations were risk factors affecting the recovery time of postoperative neurological function (P < 0.05). The area under the curve (AUC) of the Torg–Pavlov ratio, linear hyperintensity changes in the spinal cord, age at first onset, and disease duration in predicting the postoperative neurological recovery time were 0.794, 0.767, 0.772, and 0.802, respectively. The AUC predicted by the combined detection of each factor was 0.876, which was better than the area under the curve of single prediction. Conclusion Patients with CSM combined with developmental spinal stenosis were characterized by younger age of onset, a short course of the disease, and linear changes in the spinal cord high signal. The degree of developmental spinal stenosis may affect the postoperative recovery time of neurological function in CSM patients but had little effect on postoperative neurological recovery. The Torg–Pavlov ratio, age of first onset, course of the disease, and changes in the spinal cord linear hyperintensity were the factors that affected postoperative neurological recovery, which may provide a basis for reasonably predicting a postoperative neurological recovery in patients with CSM combined with developmental spinal stenosis.