Bronchial secretions of patients with cystic fibrosis (CF) inevitably become colonised with Pseudomonas aeruginosa. This organism often exhibits multiple phenotypes with different antibiotic susceptibility profiles. Isolating each colonial morphotype and determining its antibiotic susceptibility profile is labour-intensive and time-consuming. Two disk diffusion methods for mixed morphotype susceptibility testing were examined; 134 morphotypes of P. aeruginosa from 50 respiratory specimens from CF patients were tested. Mixed cultures were prepared by (a) combining morphologically different colonies of P. aeruginosa directly from the sputum specimen primary culture plates from individual patients or (b) mixing colonies of individual morphotypes after isolation and subculture. Agreement with the results for testing morphotypes individually were 85.8% and 91.6%, respectively, for the two methods. These agreement rates rose to 95.6% and 99.4%, respectively, when minor errors (intermediate misclassified as susceptible or resistant or vice versa) were excluded. Mixed morphotype testing of P. aeruginosa directly from sputum specimen culture plates in chronically colonised CF patients is more efficient, reduces turn-around time and provides clinically meaningful information about antibiotic susceptibility.