We prospectively examined 143 clinical samples from 115 patients including both HIV infected (n=53) and HIV uninfected immunocompromized (n=62) patients, with lung infiltrates and with clinical features suggestive of Pneumocystis carinii pneumonia/ PneumoCystis Pneumonia (PcP), using both microscopic techniques as well as PCR assay. Clinical samples in the present study consisted of bronchoalveolar lavage (BAL), tracheal aspirate (TA), nasopharyngeal aspirate (NPA), sputum and gastric aspirate (GA). Another group of 21 individuals with other respiratory diseases not compatible with PcP served as control during the study period of 15 months. Overall, P. jirovecii positivity rate by PCR was 12.17% (14/115 patients) compared to 3.4% (4/115) by microscopy. None of the specimens in the control group was positive by any of the techniques used. All PCR negative patients including cases and controls showed no evidence of PcP. After resolution of the discrepant results upon review of the clinical data, the sensitivity and specificity were 100% and 99%, respectively, for PCR and 30.7% and 100%, respectively, for microscopy by GMS staining. Thus, our data support the significance of PCR assay for confirming and improving the diagnosis of PcP in high-risk patients.