Extracorporeal shock wave therapy (ESWT) is seen as a therapeutic option in the treatment of chronic supraspinatus tendinitis by some authors. To test whether ESWT comprising 3 x 2000 pulses with the positive energy flux density ED+ of 0.33 mJ/mm2 is clinically superior to a sham ESWT treatment, a prospective, randomized, single-blinded, placebo-controlled study with an independent observer was performed. Forty patients were treated either by verum ESWT or sham ESWT under local anesthesia. Target criteria were the age-corrected Constant score, pain at rest and during activity on a visual analogue scale, and subjective improvement. Patients who reported no subjective improvement after 12 weeks were deblinded and received verum ESWT if they had belonged to the placebo group (partial crossover). The results of the verum group lie within the range of results for ESWT published by other authors. Patients in the placebo group with local anesthetic showed equally good results. At 12 weeks, and 1 year after intervention, no difference could be found between the verum and placebo groups regarding Constant score, pain, shoulder function, or subjective improvement. The nonresponders to the placebo ESWT continued to show no improvement after receiving verum ESWT. This contradicts a specific ESWT effect. Based on the results of this placebo-controlled study, ESWT appears to have no clinically relevant effect on supraspinatus tendinitis. The study underlines the importance of a control group in evaluating new treatment methods for diseases with unknown natural history.