INTRODUCTION: We aimed to identify the risk factors for, and outcomes of Pseudomonas aeruginosa bacteraemia in adult burns patients. METHOD: All adult burns patients who developed a Gram-negative bacteraemia over a period of 7 years were included. Retrospective data analysed included patient demographics, organisms cultured, antibiotic susceptibility patterns, isolation of P. aeruginosa in non-blood isolates, treatment, length of stay and mortality. RESULTS: Forty-three patients developed a Gram-negative bacteraemia over the study period, 12 of whom had Pseudomonas bacteraemia during the course of their admission. In eight patients (18.6%) P. aeruginosa was the first Gram-negative isolated. The only factor predicting P. aeruginosa bacteraemia as a first episode (compared to another Gram-negative) was prior isolation of Pseudomonas at other sites (wound sites, urine or sputum). Overall length of stay was less in patients who developed P. aeruginosa as a first episode, mainly because of increased mortality in this group. Prior non-blood isolates of P. aeruginosa could have correctly predicted the sensitivity pattern of the strain of P. aeruginosa organism in 75% of patients who did not receive appropriate initial antibiotics. CONCLUSION: Prior colonisation with P. aeruginosa predicts P. aeruginosa in blood cultures, as opposed to other Gram-negative bacteria. Clinicians should have a high index of suspicion for P. aeruginosa bacteraemia where a septic burns patient has a prior history of non-blood P. aeruginosa cultures. Empirical antibiotic regimes based on the antibiotic-sensitivity patterns of previous non-blood P. aeruginosa isolates in each patient should be given at the time blood cultures are taken.