Invasive pulmonary aspergillosis (IPA) is difficult to diagnose in chronic obstructive pulmonary disease (COPD). The aim of this study was to evaluate whether detection of galactomannan (GM) in bronchoalveolar lavage fluid (BALF) might be a useful means of making the diagnosis. Patients with COPD and new pulmonary infiltrates were enrolled. BALF was collected for culture and detection of GM. Venous blood was also sampled for GM detection. Biopsy samples were obtained whenever possible. Eleven cases of IPA were diagnosed (three proven and eight probable); 80 controls without IPA diagnosed were recruited. At a GM cut-off of 0.5, the sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) for diagnosing IPA were 90.9, 66.3, 27.0 and 98.1% in serum, and 90.9, 62.5, 25.0 and 98.0% in BALF, respectively. At a cut-off of 1.0, the specificity, PPV and NPV in BALF increased to 95.0, 71.4 and 98.7%; the sensitivity remained 90.9%. The sensitivity in serum was substantially lower than BALF (45.5% versus 90.9%). Receiver operating characteristic curve analysis identified an optimal BALF GM cut-off value of 1.25, with a sensitivity of 90.9% and a specificity of 96.3% for diagnosing IPA. At a relatively high cut-off value, BALF GM detection is a useful tool for the diagnosis of IPA in COPD. Besides piperacillin-tazobactam and amoxicillin-clavulanate, many other factors may also cause false-positive of GM detection in patients without IPA. Further work is needed to identify factors that might lead to false-positive or false-negative results.