Aim:In clinical practice, some patients with radicular pain do not present with typical dermatomal patterns, and selective spinal nerve block (SSNB) has been proposed to identify the involved nerve. In this study, we aimed to assess the role of our modified SSNB procedure to predict the results of the subsequent Percutaneous endoscopic transforaminal lumbar surgeries (PETLS). Material and Methods:We retrospectively analyzed data of patients who underwent our modified SSNBs before PETLS from February 2013 to March 2018 Clinical outcome data were collected 3 days after PETLS and at follow-up visits.Results:A total of 120 modified SSNB procedures (transforaminal-78 paravertebral-24, and interlaminar-18) in 92 patients presented positive response. The median follow-up period was 30.6 months. Based on Macnab criteria, the overall success rate (excellent and good results) was 83.7%. Fair and poor outcomes were observed in 10 and 5 patients, respectively. Patients with atypical extraforaminal herniations, and patients with two-level or multiple-level lumbar disc herniations or stenosis achieved desirable results after PETLS. There was significant improvement in the average VAS score for the leg three days after surgery (7.38±0.97 vs. 1.96 ±1.17, P<0.05) and on follow-up visits (1.21 ± 0.83, P<0.05). ODI was also significantly improved three days after surgery (37.20±2.36 vs. 10.95 ± 2.25, P<0.05 and at follow-up visits (8.90±1.72, P<0.05) Conclusion:The needle tip should be located closely near the intended compressed nerve via suitable approach combined with slowly injecting 1 ml lidocaine (1%) when performing our modified SSNB technique. It presents an alternative diagnostic procedure to identify the origin of pain of complicated lumbar diseases and to predict PETLS outcomes.