Intrapleural administration of fibrinolytic agents has been shown to be effective and safe in the treatment of loculated parapneumonic pleural effusions. However, controlled studies of the possible role of the activity of urokinase (UK) through the volume effect are lacking. We therefore investigated the hypothesis that UK is effective through the lysis of pleural adhesions and not through the volume effect. Thirty-one consecutive patients with multiloculated pleural effusions were randomly assigned to receive either intrapleural UK (15 patients) or normal saline (NS) (16 patients) for 3 d, in a double-blind manner. All patients had inadequate drainage through a chest tube (< 70 ml/24 h). UK was given daily through the chest tube in a dose of 100.000 IU diluted in 100 ml of NS. Controls were given the same volume of NS intrapleurally. Response was assessed by clinical outcome, fluid drainage, chest radiography, pleural ultrasonography (US) and/or computed tomography (CT). Clinical and radiographic improvement was noted in all but two patients in the UK group but in only four in the control group. The net mean volume drained during the 3-d treatment period was significantly greater in the UK group (970 +/- 75 ml versus 280 +/- 55 ml, p < 0.001). Pleural fluid drainage was complete in 13 (86.5%) patients in the UK group (two patients were treated through video-assisted thoracoscopy) but in only four (25%) in the control group. Twelve patients in the control group were subsequently treated with UK and six of them had complete drainage; the remaining six patients had complete drainage after video-assisted thoracoscopy. Our results suggest that UK is effective in the treatment of loculated pleural effusions through the lysis of pleural adhesions and not through the volume effect.