OBJECTIVES: Abacavir causes hypersensitivity reactions (HSRs) in about 4% of HIV patients. HSRs can be serious, result in hospitalisations and incur significant Health care expense. Recent evidence suggests that the presence of HLA-B57 increases the likelihood of occurrence of HSRs. The aim of this analysis is to model the cost-effectiveness of genotyping, which costs £30 per patient, as a therapeutic strategy to avoid abacavir - induced hypersensitivity reactions. METHODS: As HSRs occur within the first few weeks of therapy, a 6-month time horizon was chosen for the analysis that adopted a UK National Health Service perspective. Data relating to time of onset and prevalence of HSRs were obtained from published sources. An analysis of test characteristics (sensitivity and specificity) was performed by pooling data from two previously published studies with data of our own. The use of Health care resources, and associated costs of treating HSRs were collected from the records of 16 patients. The cost and selection of substitutes for abacavir (alternative HAART regimens), in the case of positive testing or occurrence of HSR, were considered in the analysis but assumed to be equally efficacious. A probabilistic decision analytic model (comparing testing versus no testing) was formulated and Monte Carlo simulations performed. RESULTS: When abacavir is used as salvage therapy, genotyping is cost saving. When used in treatment naïve patients or for regimen simplification, the ICER of genotyping versus conventional 'blind' therapy ranged from being cost saving to £4000 per HSR avoided, depending on the cost of alternative regimen: the more expensive the alternative, the less cost effective the test is. CONCLUSIONS: Pharmacogenetic testing to prevent abacavir - induced HSRs appears to be a cost effective use of Health care resources. A recommendation of routine adoption of testing for patients eligible for abacavir therapy may be appropriate.