BACKGROUND To assess the diagnostic accuracy and the safety of medical thoracoscopy (MT) performed with the semirigid thoracoscope. METHODS We retrospectively evaluated patients who underwent MT with semirigid thoracoscope under local anesthesia for unexplained exudative pleural effusion from March 1, 2009 to September 1, 2013 in Denmark. RESULTS Sixty-nine patients were retrospectively studied. In 13 patients it was not possible to perform the scheduled MT, in 9 cases due to an insufficient pneumothorax, in 3 due to an insufficient pleural effusion, and in 1 due to a purulent pleuritis. In 56 patients in whom MT was completed, the procedure was diagnostic in 44 cases: malignancy was reported in 26 patients and a benign diagnosis in 18. In the remaining 12 patients a definite diagnosis was not reached, and further testing was required.In an "intention-to-treat analysis" (69 patients in total), the diagnostic accuracy of MT was 63%, the sensitivity for malignancy was 59% [95% confidence interval (CI)=43%-73%], the specificity was 100% (95% CI=86%-100%), and the negative likelihood ratio was 0.41 (95% CI=0.29-0.58). Considering the 56 patients in whom it was possible to complete the procedure, the diagnostic accuracy was 78%, the sensitivity for malignancy was 74% (95% CI=54%-87%), the specificity was 100% (95% CI=83.75%-100%), and the negative likelihood ratio was 0.27 (95% CI=0.15%-0.45%). No mortality was reported. CONCLUSIONS MT performed under local anesthesia with a semirigid scope is a simple and safe procedure with an acceptable sensitivity for malignancy.