Background: Diabetes mellitus (DM) is associated with poor postoperative outcomes and increased morbidity after surgeries. Some previous studies have addressed the close association between DM and those leading to reoperations, whereas others have rejected this theory. This study aimed to evaluate the long-term effect of DM on lumbar spinal surgery using data from a nationwide sample cohort. Methods: A population-based cohort comprised one million people, which is a 2.1% representative sample of the Korean population. The present study included adult patients with lumbar degenerative diseases (e.g., lumbar spinal stenosis and spondylolisthesis), who underwent their first lumbar surgery in 2006. The cumulative incidence function for reoperation was calculated and multivariate analysis was performed to define correlation between reoperation and independent factors. Results: A total of 2020 patients were enrolled and followed up for 10 years. Nondiabetic patients, patients with DM without complication (DwoC), and patients with DM with complication (DwC) accounted for 79.5%, 9.36%, and 11.14% of all patients, respectively. Reoperation incidence stratified by DM was 12.7% for nondiabetic patients, 22.2% for patients with DwoC, and 20.0% for patients with DwC in 10 years of follow-up. During the same period, death, a competing event of reoperation, occurred in 7.8% of nondiabetic patients, in 13.2% of patients with DwoC, and in 20.9% of patients with DwC. Conclusions: DM increased 1.65 times the overall cumulative incidences of reoperation after lumbar spinal surgeries for 10 years of follow-up. The reoperation incidence for DwC may be lower than that for DwoC because of a high incidence of death as a competing event of reoperation.