To determine whether continuous application of topical glyceryl trinitrate decreases pain and symptoms in chronic noninsertional Achilles tendinopathy. Randomized double-blind placebo-controlled study of 6-months' duration. Community and referral study at an Australian University Hospital. Recruitment was through newspaper advertisements and private consulting rooms. Eligibility criteria were age >18 years, a history of insidious onset of Achilles tendon pain, a tender nodule localized to the region of the calcaneal insertion, and an ultrasound examination that excluded a tendon tear. Exclusion criteria were Achilles tendinopathy of <3 months' duration, a previous operation on, or dislocation of, the affected ankle or leg, distal neurologic signs, a local corticosteroid injection in the previous 3 months, and current pregnancy. The 65 participants (84 affected tendons; 62% men) had a median age of 49 years (range, 24-79 years), with a median duration of symptoms of 16 months (range, 4-147 months). Participants were assigned an active transdermal patch (1/4 of a Nitro-Dur 5 [Schering-Plough] glyceryl trinitrate patch), which delivered 1.25 mg of glyceryl trinitrate over 24 hours, or a placebo patch. Patients were required to cut the patches into quarters and apply 1/d to the site of maximal tenderness for the 24-week duration of the study. All patients were also given 500-mg paracetamol tablets for use with headaches, and instructed in a rehabilitation program that comprised rest from aggravating activities, the use of heel-raise wedges, prolonged daily stretching of the gastrocnemius and soleus musculature, and an eccentric calf muscle-strengthening program. At the baseline, 2, 6, 12, and 24-week examinations the patient completed a symptom assessment sheet to rate the severity of Achilles pain with activity, at rest, and at night (0 = no pain, 4 = very severe pain). The single assessor used the same scale to measure local tenderness; an 11-point scale for the patient to report pain after the single-leg 10-hop test; and also measured the ankle plantar flexor mean peak force and ankle plantar flexor work. Follow-up was 89% complete. The groups did not differ in pain with activity, night pain, or local tenderness until the 12-week assessment when participants in the glyceryl trinitrate group reported less pain on each measure (mean scores, 0.9 vs. 1.6 [P = 0.02]; 0.2 vs. 0.7 [P = 0.04]; and 0.9 vs. 1.6 [P = 0.02], respectively). The difference was maintained at 24 weeks for pain with activity (mean scores, 0.4 vs. 1.0 [P = 0.03]). At 24 weeks the glyceryl trinitrate group reported less pain on the 10-hop test than the placebo group (mean scores, 0.5 vs. 1.6 [P = 0.005]). Although the intervention group showed a greater increase in plantar flexor mean total work at 24 weeks than the placebo group, the baseline scores were significantly different. The groups did not differ in pain at rest or in ankle plantar flexor peak force. Combining all the measures showed an estimated 14% (95% CI, 9%-19%) excess of asymptomatic tendons in the intervention group at 6 months. Reported main side effects were headaches (glyceryl trinitrate group, 53%; placebo group, 45%) and rashes (glyceryl trinitrate group, 16%%; placebo group, 12%). A topical glyceryl trinitrate patch was more effective than placebo for reducing pain from chronic noninsertional Achilles tendonitis in the first 12 and 24 weeks of use.