Genome-wide association studies (GWAS) on warfarin and acenocoumarol showed that interindividual dosage variation is mainly associated with single nucleotide polymorphisms (SNPs) in VKORC1 and to a lesser extent in CYP2C9 and CYP4F2. For phenprocoumon dosage, the genes encoding CYP3A4 and ApoE might play a role.To assess the association between common genetic variants within VKORC1, CYP2C9, CYP4F2, CYP3A4, and ApoE and phenprocoumon maintenance dosage, and to identify novel signals using GWAS.We selected all participants from the Rotterdam study who were treated with phenprocoumon. For each SNP, we tested the association between the above-mentioned genotypes and age, sex, body mass index, and target INR adjusted-phenprocoumon maintenance dosage.Within our study population (N=244), VKORC1, CYP2C9, CYP4F2 genotypes together explained 46% of phenprocoumon maintenance dosage variation. Each additional VKORC1 variant allele reduced phenprocoumon maintenance dosage by 4.8 mg/week (P<0.0001) and each additional CYP2C9 variant allele by 2.2 mg/week (P=0.002). Each additional variant allele of CYP4F2 increased phenprocoumon dosage by 1.5 mg/week (P=0.022). Variant alleles of CYP3A41*B and ApoE showed no association with phenprocoumon dosage. Genome-wide significant SNPs were all related to VKORC1 activity. Best associated were two SNPs in complete linkage disequilibrium with each other and with SNPs within VKORC1: rs10871454 [Syntaxin 4A (STX4A)] and rs11150604 (ZNF646), each with a P value of 2.1×10⁻²². Each reduced phenprocoumon maintenance dosage weekly by 4.9 mg per variant allele.Similar to earlier findings with warfarin and acenocoumarol, phenprocoumon maintenance dosage depended on polymorphisms in the VKORC1 gene. CYP2C9 and CYP4F2 were of modest relevance.