Genetic factors importantly contribute to migraine. However, unlike for rare monogenic forms of migraine, approaches to identify genes for common forms of migraine have been of limited success. Candidate gene association studies were often negative and positive results were often not replicated or replication failed. Further, the significance of positive results from linkage studies remains unclear owing to the inability to pinpoint the genes under the peaks that may be involved in migraine. Problems hampering these studies include limited sample sizes, methods of migraine ascertainment, and the heterogeneous clinical phenotype. Three genome-wide association studies are available now and have successfully identified four new genetic variants associated with migraine. One new variant (rs1835740) modulates glutamate homeostasis, thus integrates well with current concepts of neurotransmitter disturbances. This variant may be more specific for severe forms of migraine such as migraine with aura than migraine without aura. Another variant (rs11172113) implicates the lipoprotein receptor LRP1, which may interact with neuronal glutamate receptors, thus also providing a link to the glutamate pathway. In contrast, rs10166942 is in close proximity to TRPM8, which codes for a cold and pain sensor. For the first time this links a gene explicitly implicated in pain related pathways to migraine. The potential function of the fourth variant rs2651899 (PRDM16) in migraine is unclear. All these variants only confer a small to moderate change in risk for migraine, which concurs with migraine being a heterogeneous disorder. Ongoing large international collaborations will likely identify additional gene variants for migraine.