OBJECTIVES: The aim of the study was to identify possible causes of pancreatic insufficiency in patients with HIV infection. METHODS: A retrospective analysis of 233 HIV-positive patients for whom faecal elastase measurement was available was performed to investigate potential associations with core demographic data, HIV infection characteristics, degree of immunosuppresion, exposure to antiretroviral therapy (ART), alcohol misuse, diabetes, hepatitis C virus (HCV) infection, triglyceride and cholesterol levels and symptomatology. The response to pancreatic enzyme replacement for patients with evidence of insufficiency was also evaluated. RESULTS: Of 233 patients, 104 (45%) had evidence of pancreatic exocrine insufficiency (faecal elastase < 200 mcg/g). A positive association with exocrine pancreatic insufficiency was found for HCV infection (P = 0.007), previous or current HCV treatment (P = 0.003), alcohol misuse history (P = 0.006) and the presence of steatorrhoea (P = 0.03). There was no demonstrated association between exocrine pancreatic insufficiency and didanosine (ddI) exposure (P = 0.43) or stavudine (d4T) exposure (P = 0.62). Seventy-seven per cent of patients who were treated with pancreatic enzymatic supplementation reported a subjective improvement in symptoms. CONCLUSIONS: Faecal elastase sampling should form part of the routine work-up for HIV-positive patients with chronic diarrhoea even in the absence of 'traditional' risk factors such as ddI exposure. In particular, if the patient has steatorrhoea, a history of alcohol exposure or their HCV serology is positive, they should be considered for investigation. Treatment with pancreatic enzyme supplementation appears to be effective in the treatment of chronic diarrhoea caused by pancreatic insufficiency in the majority of patients.