BACKGROUND: Tuberculous pleurisy has been treated with a combined regimen of corticosteroids-antimicrobial therapy. But whether such combination therapy add to benefits is unknown. We evaluate the effects of corticosteroid and its routine application in relief of clinical symptoms, absorption of pleural effusion, and pleural adhesions. METHODS: A prospective, randomized study of the role of corticosteroid in the treatment of tuberculous pleurisy was performed in 83 patients(nonsteroid group:50 patients, steroid group: 33 patients) from June, 1991 to September, 1994. RESULTS: 1) The mean duration from symptoms(fever, chest pain, dyspnea) to relief was 3.8 days in the steroid group, and 7.4 days in the nonsteroid group(P<0.05). Clinical symptoms including fever, chest pain, sputum and weight loss were relieved more rapidly in the steroid group than other symptoms(weakness, night sweating and dyspnea). 2) Pleural effusion was taken an averge of 88 days in the steroid group and 101 days in the nonsteroid group to be absorbed completely(p>0.05). 3) The incidence of pleural adhesions was 17/33(51.5%) in the steroid group and 32/50(64%) in the nonsteroid group(p>0.05). 4) Side effects of corticosteroids were observed in only one patient causing epigastric pain and discontinuation of drug. CONCLUSION: Corticosteroid exert benefitial role in the more rapid relief of clinical symptoms to patients with tuberculous pleurisy, but absortion of pleural effusion and occurrence of pleural adhesions was not influenced significantly. Therefore, its routine application should be reevaluated.