The purpose of this study was to evaluate the efficacy, safety, and cost of bedside pleurodesis for malignant pleural effusions using talc slurry (TS) or bleomycin (BL) in a prospective randomized trial, and to determine prognosticators for procedure failure. From June 1997 to June 1999 a series of 71 patients entered this trial. They underwent 37 procedures with TS (4 g) and 34 with BL (60 units) via tube thoracostomy. Success was defined as no recurrence of pleural effusion or asymptomatic recurrence of a small amount of effusion. Pleural effusion-free survival curves were used to analyze the success rates and the prognosticators of failure. Follow-up ranged from 3 days to 26 months (median 2.5 months). No difference in success rates was detected between TS or BL (log-rank test: p = 0.724). There were no major complications related to the procedure. The independent prognosticators of failed pleurodesis were the use of steroids ( p = 0.004) and the volume of pleural fluid drained during the first thoracentesis when it was more than 900 ml ( p = 0.029). The average cost of intervention per patient was significantly lower for TS ( p < 0.001). There was no significant difference between the success rates for TS and BL as agents of bedside pleurodesis for malignant pleural effusions. Because of its significantly lower cost, TS should be considered the agent of choice. The use of steroids and the volume drained during the first thoracentesis (if more than 900 ml) were independent prognosticators of pleurodesis failure. The role of this latter finding as a marker of pleurodesis failure awaits more data.