The most important primary headaches (i.e. independent disorders that are not caused by another disease) are migraine, tension-type headache and cluster headache. All primary headaches are in need of better treatments. Migraine has a prevalence of 10% in the general population and its societal costs are high. Although the precise mechanisms underlying the pathophysiology of migraine are still elusive, the last decades have witnessed some progress (e.g. involvement of serotonin, calcitonin gene-related peptide, nitric oxide, etc). Nitric oxide (NO) is a very important molecule in the regulation of cerebral and extra cerebral cranial blood flow and arterial diameters. It is also involved in nociceptive processing. Glyceryl trinitrate (GTN), a pro-drug for NO, causes headache in normal volunteers and a so called delayed headache that fulfils criteria for migraine without aura in migraine sufferers. Blockade of nitric oxide synthases (NOS) by L-NMMA effectively treats attacks of migraine without aura. Similar results have been obtained for chronic tension-type headache and cluster headache. Inhibition of the breakdown of cGMP also provokes migraine in sufferers, indicating that cGMP is the effector of NO-induced migraine. Several relationships exist between NO, calcitonin gene-related peptide and other molecules important in migraine. Also ion channels, particularly the K(ATP) channels, are important for the action of NO. In conclusion, inhibition of NO production or blockade of steps in the NO-cGMP pathway or scavenging of NO may be targets for new drugs for treating migraine and other headaches. Indeed, selective n-NOS and i-NOS inhibitors are already in early clinical development.