Objectives: This study analyzed the diagnostic and therapeutic usefulness of selective nerve root block (SNRB) in patients with spinal stenosis who show inconsistencies between magnetic resonance imaging (MRI) and clinical findings (e.g., dermatomes).Summary of Literature Review: MRI is known to be highly accurate for diagnosing spinal stenosis, but few studies have investigated the diagnostic accuracy of MRI for spinal canal stenosis. In addition, the most suitable treatment of patients with inconsistent clinical and MRI findings has not been established. Methods: This single-center, retrospective cohort study was conducted among 93 patients treated between January 2013 and December 2018, who underwent at least two SNRBs for single-level spinal stenosis on MRI with clinical discrepancies. Seventeen patients who were diagnosed with other causes of pain (e.g., spondylolisthesis, sacroiliac joint dysfunction, lower leg arterial occlusion) were excluded. The first SNRB targeted the lesion found on MRI. One week later, patients were assessed using a visual analog scale (VAS) and a second procedure was performed on the dermatome-indicated level if there was no improvement. VAS scores were also obtained 3 months after the procedure. Results: In total, 45 patients had central stenosis and 31 patients had foraminal stenosis. The average VAS score before the SNRB was 7.4±1.4. After the first procedure, the average VAS score was 5.2±2.6. After the second procedure, the average VAS score was 2.4±1.3, and after 3 months, it was 3.6±1.9. Of the patients with foraminal stenosis, 77.4% did not respond to the first block (based on MRI), but responded well to the second procedure (based on clinical symptoms) (p<0.05). Conclusions SNRB targeting the level corresponding to clinical symptoms may be useful for locating the symptomatic nerve root and providing pain relief in patients with foraminal stenosis on MRI with discordant clinical findings.