Tuberculous pleural effusion (TPE) is a common problem for differential diagnosis from malignant effusion (MPE) in epidemic areas of tuberculosis (TB). Prediction based on adenosine deaminase (ADA) is dependent on age as well as the tuberculosis incidence. The aim of the study was to evaluate cutoff values for ADA with sensitivity and specificity results for the differential diagnosis of MPE and TPE in a population with intermediate incidence of TB. We retrospectively analysed 196 patients with a definitive diagnosis of TPE (n = 114) and MPE (n = 82). The optimal cutoff value of ADA was determined using the receiver operating characteristic (ROC) curve. There was a statistically significant difference according to the levels of pleural fluid ADA between TPE and MPE groups (p < 0.0001). The cutoff value for diagnosing TPE was > 55 U/L, with a sensitivity = 86.8%, specificity = 86.6%, positive predictive value (PPV) = 90%, negative predictive value (NPV) = 82.6% and accuracy = 82.6%. We then combined ADA > 55 U/L and age < 50 and were able to discriminate the TPE group with increased specifity (95.7 %) and PPV (98.8%) results. The model could correctly classify 21 MPE out of 23 and 82 TPE out of 94 patients. A pleural fluid ADA value < 31 U/L suggests that TPE is highly unlikely with a sensitivity = 43.9 %, specificity = 100%, PPV = 100%, NPV = 71.3% and accuracy = 76.6%. It can be concluded that ADA is a very useful parameter for the differential diagnosis of TPE and MPE, specifically in youngers with a higher incidence of tuberculosis.