Background: There are numerous radiological and anatomical studies on lumbar foramina in the literature, but there are no distinctive studies about the relationship between treatment and the type of foraminal stenosis. This study was conducted to better evaluate foraminal stenosis and to plan treatment accordingly.Methods: Foraminal stenosis was divided into 2 groups: stable and unstable stenosis. Both groups were also divided into 4 subgroups in relation to the cause and type of compression and based on the structure of the intervertebral disc. The visual analog scale for leg pain (VAS-LP) and Oswestry Disability Index (ODI) scores were investigated before and after surgery.Results: A total of 115 patients (59 women and 56 men) underwent surgery for lumbar foraminal stenosis. The mean patient age was 56.1 years (range 17-80 years). The mean follow-up was 29 months (range 24-39 months). There were 36 patients (32%) with stable foraminal stenosis and 79 patients (68%) with unstable foraminal stenosis. The majority of the patients were identified as having unstable type 1 foraminal stenosis (45 of 115). The VAS-LP and ODI scores for each group decreased gradually during the follow-up periods and showed significant decrease during the last follow-up (P < 0.001). Interobserver and intraobserver agreement in the classification of foraminal stenosis was found to be nearly perfect. No patients experienced postoperative radiculopathy complication. Only 2 patients experienced superficial operation site infection and 1 showed deep wound infection. The patient who had a deep wound infection needed to repeat surgery for the infection.Conclusions: We introduced a novel classification system for lumbar foraminal stenosis. We aimed to guide appropriate treatment modality depending on the determined classification. This classification helps to determine the optimal treatment. In the light of our findings, the patients who were operated according to our classification experienced satisfactory clinical outcomes and low complication rates.