Since its inception in the early 1990s, arterial spin labelling (ASL) has been developed in numerous variants. Tremendous improvements in both pulse sequences and hardware have allowed improved image quality and reduced scan time, to the point where the technique might be ready for clinical application. However, although now routinely used in research centres and universities, its application in daily clinical routine remains restricted to a few centres. Its popularity in magnetic resonance imaging (MRI) research centres and among physicists may have contributed to the lack of enthusiasm from the clinical community, as the variety of possible sequences available might deter even the most die-hard technology-driven neuroradiologist from using this technique. Therefore, a joint action directed towards the harmonisation of all the existing ASL pulse sequences and the development of a reliable and common post-processing pipeline to provide guidance for the major MRI manufacturers was recently endeavoured and dubbed the 'ASL Network' ( http://www.asl-network.org ). This network was established to provide a communication platform among physicists, engineers and clinicians who are keen to see this technique finally make it to daily clinical practice. Finally, a networking European Grant from the COST Office was secured to help fund some of the activities of the ASL Network, in particular as they pertain to the application of ASL in dementia ( http://www.cost.eu/domains_actions/bmbs/Actions/BM1103 ). Here, both the ASL Network and the COST Action are described, with their respective goals and aspirations.