Background: EUS has been shown to be effective in detecting small amounts of ascites, which are frequently not detected by computed tomography (CT). EUS imaging of the mediastinum can also detect pleural fluid (Gastro Endo 1999). No prior series report the sensitivity of EUS in detecting pleural fluid or EUS-guided FNA in performing thoracentesis. Aim: To determine the sensitivity of EUS when compared to CT scanning and to evaluate the utility of EUS-guided FNA for pleural fluid. Methods: Between the periods of May 1994 through October 1999, 1048 patients underwent upper EUS exams for various indications. Retrospectively, 130 patients found to have pleural effusions were evaluated using File Maker ProTM software, chart review and telephone interviews. EUS was performed using the Olympus echoendoscope (GF-UM20/130) and EUS-guided FNA with the Pentax echoenodscope (FG32UA/36UX) or Olympus echoendoscope (GF-UC30P) in conjunction with the GIP-Mediglobe or Wilson Cook 22-gauge needle. Fluid was seen and aspirated from the right pleural space, via a trans-esophageal approach. As patients are in the left lateral decubitus position, only right pleural effusions can be seen near the esophagus. Results: Of the 130 patients found to have a pleural effusion by EUS, CT scans were performed before EUS in 89. Prior to EUS, 35 of 89 (40%) of patients had CT scans which did not detect a pleural effusion. Of the 130 patients, 20 (15%) patients had EUS-guided FNA of the pleural effusion detected. The mean volume of fluid obtained was 6cc (range 1- 24cc). The average number of passes required to obtain fluid was 1.05 passes. Three patients had a diagnosis of malignant pleural effusion on EUSguided FNA. In two of the patients with malignant pleural effusions, CT failed to detect the pleural fluid. The diagnosis of metastatic disease in these patients precluded surgery. On follow up of the 20 patients who had EUS-guided thoracentesis, 16 patients were found to have a final diagnosis of cancer. Of these 20 patients, 11 did not have a previously detected pleural effusion. In these 11 patients, 3 subsequently developed a pleural effusion. No complications occurred after EUS-guided FNA of pleural fluid. Conclusions: EUS is a sensitive method of detecting small amounts of pleural fluid, which are often not seen by CT scans. EUS-guided thoracentesis is feasible and it can establish the diagnosis of metastatic disease, altering patient management. Background: EUS has been shown to be effective in detecting small amounts of ascites, which are frequently not detected by computed tomography (CT). EUS imaging of the mediastinum can also detect pleural fluid (Gastro Endo 1999). No prior series report the sensitivity of EUS in detecting pleural fluid or EUS-guided FNA in performing thoracentesis. Aim: To determine the sensitivity of EUS when compared to CT scanning and to evaluate the utility of EUS-guided FNA for pleural fluid. Methods: Between the periods of May 1994 through October 1999, 1048 patients underwent upper EUS exams for various indications. Retrospectively, 130 patients found to have pleural effusions were evaluated using File Maker ProTM software, chart review and telephone interviews. EUS was performed using the Olympus echoendoscope (GF-UM20/130) and EUS-guided FNA with the Pentax echoenodscope (FG32UA/36UX) or Olympus echoendoscope (GF-UC30P) in conjunction with the GIP-Mediglobe or Wilson Cook 22-gauge needle. Fluid was seen and aspirated from the right pleural space, via a trans-esophageal approach. As patients are in the left lateral decubitus position, only right pleural effusions can be seen near the esophagus. Results: Of the 130 patients found to have a pleural effusion by EUS, CT scans were performed before EUS in 89. Prior to EUS, 35 of 89 (40%) of patients had CT scans which did not detect a pleural effusion. Of the 130 patients, 20 (15%) patients had EUS-guided FNA of the pleural effusion detected. The mean volume of fluid obtained was 6cc (range 1- 24cc). The average number of passes required to obtain fluid was 1.05 passes. Three patients had a diagnosis of malignant pleural effusion on EUSguided FNA. In two of the patients with malignant pleural effusions, CT failed to detect the pleural fluid. The diagnosis of metastatic disease in these patients precluded surgery. On follow up of the 20 patients who had EUS-guided thoracentesis, 16 patients were found to have a final diagnosis of cancer. Of these 20 patients, 11 did not have a previously detected pleural effusion. In these 11 patients, 3 subsequently developed a pleural effusion. No complications occurred after EUS-guided FNA of pleural fluid. Conclusions: EUS is a sensitive method of detecting small amounts of pleural fluid, which are often not seen by CT scans. EUS-guided thoracentesis is feasible and it can establish the diagnosis of metastatic disease, altering patient management.