OBJECTIVE: To test a case-definition score for its usefulness in the diagnosis of pulmonary aspergillosis and to determine microbiological criteria useful in assessing the significance of isolating Aspergillus species from lower respiratory tract specimens (LRTS). METHODS: With the exception of cystic fibrosis patients, all patients with Aspergillus isolated from LRTS were classified according to a clinical case-definition ratio as: (i) colonized patients; (ii) patients with isolates of undetermined significance; (iii) patients with proven or probable pulmonary aspergillosis. Retrospective review of pathology records led to the identification of additional patients with histologically compatible pulmonary aspergillosis. Microbiological variables were compared between aspergillosis and colonization groups. RESULTS: Seventy-six isolates of Aspergillus species were obtained from 73 patients. For patients with proven (n=2) or histologically compatible pulmonary aspergillosis (n=5), the case definition ratio was always higher than 0.3; no patient with proven disease was classified into the colonized group. Compared to patients considered to be colonized (n=26), patients with proven or probable pulmonary aspergillosis showed a significant difference in the total number of Aspergillus colonies cultured per episode. In addition, for bronchoalveolar lavage specimens, the mean number of Aspergillus colonies/agar plate was also significantly higher in the latter group. The sensitivity and specificity of Aspergillus spp. detection by culture of LRTS for proven or probable pulmonary aspergillosis was 35.7% and 70.4%, respectively. CONCLUSION: The case-definition ratio used in this study helps interpret the clinical significance of Aspergillus isolated from LRTS. Data for assessment should include the absolute number of colonies recovered from culture of any LRTS and, for bronchoalveolar lavage, the mean number of Aspergillus colonies per agar plate.