Many studies have been reported on recurrent lumbar disc herniations covering several pathological conditions. In those studies, reoperation rate of revised disc excisions was calculated by simple division between the number of reoperations and that of the total primary disc excisions. To determine the real reoperation rate, strict definition of pathologies, a large number of patients, a long observation period, and survival function method are necessary. Methods Between 1988 and 2007, 5,626 patients with disc excision were enrolled by the spine registration system of the Department of Orthopaedic Surgery, Tohoku University, Japan. Among them, 192 had revised disc surgery, and we obtained data of 186 patients whose clinical features were assessed and reoperation rates analyzed using the Kaplan-Meier method. Results In total, 205 disc herniations were excised in the revision surgery (including contralateral herniation at the same level and new herniation at a different level), and 101 were real recurrent herniations (recurrence at the same level and side as the primary herniation). The kappa coefficient of the spinal level and side between the primary and revision surgeries was 0.41, indicting moderate correlations. Real recurrent herniations showed shorter intervals between primary and revision surgeries. Male patients with surgery at a younger age carried a higher risk of reoperation. In the revision surgery, transligamentous extrusion was significantly more common than other types of herniation. On Kaplan-Meier analysis, the reoperation rate of overall revised excisions was 0.62% at 1 year, 2.4% at 5 years, 4.4% at 10 years, and 5.9% after 17 years. That of real recurrent herniations was 0.5%, 1.4%, and 2.1%, respectively, and 2.8% after 15.7 years. Conclusion Reoperation rate of real recurrent herniations calculated using survival function method gradually increased year by year, from 0.5% at 1 year after primary surgery to 2.8% at 15.7 years.