This randomized controlled study addressed whether sonographic needle guidance affected the outcomes of corticosteroid injection for symptomatic carpal tunnel syndrome. Seventy-seven symptomatic carpal tunnels were randomized to injection by either (1) conventional anatomic landmark palpation-guided injection or (2) sonographic image-guided injection, each using a two-step technique where 3 ml of 1% lidocaine was first injected to hydrodissect and hydrodisplace critical intra-carpal tunnel structures followed by injection with 80 mg of triamcinolone acetonide (2 ml). Baseline pain, procedural pain, pain at outcome (2 weeks and 6 months), responders, therapeutic duration, total cost, and cost per responder were determined. There were no complications in either treatment group. Relative to conventional anatomic landmark palpation-guided methods, sonographic guidance for injection of the carpal tunnel resulted in 77.1% reduction in injection pain (p<0.01), a 63.3% reduction in pain scores at outcome (p<0.014), 93.5% increase in the responder rate (p<0.001), 84.6% reduction in the non-responder rate (p<0.001), a 71.0% increase in therapeutic duration (p<0.001), and a 59.3% ($150) reduction in cost/responder/year for a hospital outpatient (p<0.001). However, despite improved outcomes, cost per patient per year was significantly increased for an outpatient in a physician's office and was neutral for a hospital outpatient. Sonographic needle guidance significantly improves the performance and clinical outcomes of injection of the carpal tunnel and is cost-effective for a hospital-based practice, but based on current reimbursements, it significantly increases overall costs for medical care delivered in a non-hospital-based physician practice.