OBJECTIVES: Carbamazepine (CBZ) is a widely-used, first-line treatment in epilepsy. However, CBZ is associated with hypersensitivity adverse drug reactions (ADRs) ranging from mild rash such as maculopapular exanthema, to hypersensitivity syndrome, Steven-Johnson syndrome and toxic epidermal necrolysis (TEN). TEN is associated with a mortality rate of up to 30%. The presence of the HLA-A*3101 allele is associated with an increased risk of CBZ-induced hypersensitivity reactions [OR 9.1, 95% CI, 4.0 to 20.7]. HLA-A*3101 is present in 2% - 5% of populations of Northern European descent. We aim to investigate the cost effectiveness of pharmacogenetic testing for HLA-A*3101prior to initiation of CBZ treatment in patient with epilepsy. Patients testing positive for the allele are prescribed an alternative antiepileptic drug, lamotrigine. METHODS: A decision analytic model was developed to represent the first three months post initiation of anti-epileptic drug, to cover the period when the majority of severe ADRs manifest. A Markov model (cycle length 1 year) was used to simulate costs (from the perspective of the NHS in the UK) and utilities incurred in subsequent years. This enables modelling of costs and disutilities from long term sequelae of severe ADRs as well as the effectiveness of treatment in terms of remission of seizures. Transition probabilities, costs and utilities were sourced from patient level data from the SANAD trial [Lancet 369(9566):1000-15] and relevant literature. RESULTS: Compared with no pharmacogenetic testing, and prescribing CBZ for all patients, the test results in an incremental cost effectiveness ratio of £26,684 per QALY gained. The probability that testing is cost-effective at a threshold of £30,000 per QALY is 0.55, and the cost of preventing a single ADR is £35962. CONCLUSIONS: Pharmacogenetic testing for HLA-A*3101 prior to treatment with CBZ might be cost-effective for populations of North European descent.