Objectives: The primary objective of this study was to describe the accuracy of pneumonia diagnosis, both community-acquired pneumonia (CAP) and hospitalacquired pneumonia (HAP). Secondary objectives were describing the choice of antibiotics used, pathogens isolated, and predictive parameters in diagnosing pneumonia. Methods: This was a prospective cross-sectional study to determine the accuracy of the diagnosis of CAP and HAP admitted to Hospital Tuanku Ja’afar. All patients aged ≥12 years admitted to the general medical ward with the diagnosis of CAP or HAP were included in the study. Chest radiograph interpretation was done by certified radiologists. An accurate diagnosis of pneumonia was defined by clinical signs and symptoms of pneumonia supported by radiographical evidence. Results: A total of 159 patients were enrolled into the study from January 2018 to February 2018. Of these only 59(37.1%) cases were accurately diagnosed as pneumonia. Amongst those with pneumonia diagnosis made by the emergency department, medical officers and specialists of medical department; 65.4%, 60% and 47.3% respectively were not pneumonia. Amoxicillin with clavulanate and azithromycin were amongst the most common first choice of antibiotic used (46.5%). In this study, pathogens were isolated either by blood culture or sputum culture in only 20 (12.6%) patients. There was no significant predictive parameter identified in this study, which included white cell counts, Creactive protein (CRP) levels, erythrocyte sedimentation rate (ESR), and Pao2/FiO2 ratio. Conclusion: About two-thirds of patients diagnosed with pneumonia did not have a compatible radiological finding. Better tools and systems are needed to aid in the diagnosis of pneumonia