STUDY DESIGN: Prospective, masked, double controlled diagnostic trial. OBJECTIVES: To determine the sensitivity and specificity of electrodiagnostic consultation (EDX) for the clinical syndrome of lumbar spinal stenosis. SUMMARY OF BACKGROUND DATA: EDX has been used for more than 50 years to diagnose spinal disorders but has not met the new standards of evidence-based medicine. METHODS: A total of 150 subjects (asymptomatic volunteers and patients with MRIs suggesting back pain or spinal stenosis; 55-80 years of age) underwent physiatrist history and physical examination, MRI, and review of this data by a neurosurgeon, with each clinician masked to any outside information, leading to a unanimous consensus on diagnosis in 55. After masked EDX testing, 7 subjects with undiagnosed neuromuscular disease were discovered. EDX findings were related to "clinical gold standard" diagnoses in 48 persons. RESULTS: Paraspinal mapping EMG score of >4 had 100% specificity and 30% sensitivity for stenosis compared with either the back pain or asymptomatic groups (each, P < 0.04). A composite limb and paraspinal fibrillation score had a sensitivity of 47.8% and specificity of 87.5% (P = 0.008), and H-wave sensitivity was 36.4, specificity 91.3 (P = 0.026) for stenosis versus all controls. CONCLUSIONS: This first masked study in the 60-year history of needle electromyography also introduces anatomically validated needle placement, quantified and reproducible examination of the paraspinal muscles, and dual control populations to EDX research in spinal disorders. EDX has statistically significant, clinically meaningful specificity for spinal stenosis and detects neuromuscular diseases that may masquerade as stenosis.