The outcomes of surgical treatment in 80 patients with cervical compressive myelopathy were retrospectively reviewed to examined the correlations between surgical outcomes and the following seven predictive factors: age at surgery, duration of symptoms, severity of myelopathy, number of compressed segments, intramedullary high intensity segments on T₂-weighted magnetic resonance (MR) imaging, surgical method, and the type of disease. The recovery rates were evaluated at 3 months after the surgery. Significant correlations were observed between recovery rate and duration of symptoms, severity of myelopathy, and high intensity segments on T₂-weighted MR imaging. No statistical correlation was observed with the other factors. Multivariate analysis revealed significant correlations between recovery rate and duration of symptoms and number of high intensity segments on T₂-weighted MR imaging. The multiple regression equation was expressed as follows: recovery rate = 82.981 + 0.101 × (age) - 0.675 × (duration) - 1.452 × (number of compressed segments) - 1.451 × (preoperative Neurosurgical Cervical Spine Scale) - 13.826 × (number of high intensity segments). Based on this predicted formula, we compared the predicted and actual recovery rates for 17 patients treated recently. The two values were similar except in two patients with long duration of symptoms. We conclude that the surgical outcome can be predicted to a certain extent and this information could be provided to patients considering surgery for cervical compressive myelopathy.