To examine whether diaphragmatic plication is an effective and lasting treatment option for non-malignant diaphragmatic paralysis. Nineteen patients who had undergone diaphragm plication (1983-1990) were recalled for interview, pulmonary function testing and chest X-ray. There were 13 men and six women aged 24-73 (mean 55). Diaphragm paralysis was idiopathic (n=9), postsurgical (n=3), related to cervical spondylosis (n=4) and neck injury (n=2). Patients presented with breathlessness (18/19) or orthopnoea (1/19). Symptoms had lasted 3-60 months (mean 24 months). All patients had a raised hemidiaphragm on chest X-ray with paradoxical movement on ultrasound. Mean preoperative FVC was 71% predicted (range 38-93, SD 12.9) and mean FEV(1) was 67% predicted (range 33-90, SD 10.8). Supine lung volumes were 81% (mean) of sitting values. There were six right plications and 13 left. There were no postoperative deaths. One patient required re-plication. Follow-up (18/19 of original operated patients) ranged from 7-14 years (mean 10 years). Three patients had died of unrelated causes and one patient failed to attend long term follow-up, leaving 15 patients of the original 19 operated on. Positional change in lung volumes was not affected by surgery at early (6 week) or late (>5 year) follow-up. FVC, FEV(1), FRC and TLC improved by 10.1*, 11.8*, 16.9* and 9.2*%, respectively, at early follow-up and 11.8*, 15.4*, 26 and 13.3*% at late follow-up (*P<0.005 signed rank). Dyspnoea scores at long term follow-up improved 1 point (n=5), 2 points (n=5) and 3 points (n=2), remained unchanged (n=1) or dropped 1 point (n=2). Of the 15 patients followed up all but one who had been employed returned to work. 14/15 patients expressed satisfaction with their surgery. Diaphragm plication is an effective procedure with lasting results.