Background: More than 20% of inflammatory bowel diseases (IBD) patients discontinue thiopurine therapy due to severe adverse drug reactions among which leucopenia is one of the most serious. Thiopurine S-methyltransferase (TPMT) pharmacogenetics has been proven effective for optimizing azathioprine/mercaptopurine safety. Nevertheless, TPMT screening is used in clinical practice on a very limited scale. The aim of our study was to assess whether pre-treatment TPMT genotype-based dosing reduces the occurrence of leucopenia and whether this strategy is cost-effective. Methods: We performed a randomized trial in thiopurine naïve IBD patients starting on thiopurine treatment [the Dutch “Thiopurine response Optimization by Pharmacogenetic testing in Inflammatory bowel disease Clinics” (TOPIC) study (ClinicalTrials.gov: NCT00521950)]. Patients were randomly assigned to pre-treatment screening for three common variants in TPMT (TPMT∗2, ∗3A and ∗3C) or standard thiopurine treatment. Patients heterozygous for a TPMT variant received 50% of the standard thiopurine dose, patients homozygous for the tested variants 0-10%. We compared pre-treatment genotyped patients with patients receiving standard dose for the occurrence of leucopenia (leucocyte count