Introduction.Burkholderia pseudomallei (melioidosis) is an important cause of community-acquired pneumonia (CAP) in the tropics. Selective medium is recommended for laboratory diagnosis with non-sterile respiratory samples, while PCR is not routinely used due to variable reported performance. The effectiveness of these diagnostic modalities varies by site.Aim. To compare selective media and real-time PCR (qPCR) with routine media in detecting B. pseudomallei in CAP respiratory samples in a low-incidence setting in Kuala Lumpur, Malaysia.Methodology. Respiratory samples were routinely cultured on blood, chocolate and MacConkey agar (RESP-ROUTINE), and compared to culture on selective Ashdown medium (RESP-SELECTIVE) and qPCR. The gold standard was routine culture of B. pseudomallei from any site (ALL-ROUTINE).Results.B. pseudomallei was detected in 8/204 (3.9 %) samples. Overall sensitivity rates differed (P=0.03) for qPCR (100%), RESP-SELECTIVE (87.5%) and RESP-ROUTINE (50%). There was a trend towards lower median days to positive culture for RESP-SELECTIVE (1 day) compared to RESP-ROUTINE (2 days, P=0.08) and ALL-ROUTINE (2 days, P=0.06). Reagent costs for each additional detection were USD59 for RESP-SELECTIVE and USD354 for PCR.Conclusions. In a low-incidence setting, selective culture of respiratory samples on Ashdown was more sensitive and allowed quicker identification than routine media, at reasonable cost. Blood cultures are critical, confirming four cases missed by routine respiratory culture. Selective medium is useful in early pneumonia (pre-sepsis) and resource-limited settings where blood cultures are infrequently done. Real-time PCR is costly, but highly sensitive and useful for high-risk patients with diabetes, cancer or immunosuppressants, or requiring ventilation or intensive care.