Over the past 10 years, research involving the mammalian blood vessel (my 'old friend') has involved major surprises. 1. The realisation that the internal elastic lamina is full of holes, through which endothelial cells send feet-like projections, has highlighted the diffusion pathways that are surely involved in the modulation of myocyte activity by endothelium-derived factors. 2. The roles of the Ca2+-sensitive K+ channels, SKCa and IKCa, present on the vascular endothelium is not totally clear but the demonstration of their localisation in different endothelial microdomains has been a fundamental advance. 3. The discovery that vascular endothelial cells possess a G protein-coupled receptor (the CaS) which senses extracellular Ca2+ was not expected. Even more surprising is the close coupling between the CaS and IKCa channels, with activation of the former triggering opening of IKCa without any involvement of SKCa. Speculatively, the function of the vascular CaS is to respond to increases in [Ca2+] in myoendothelial spaces (especially those associated with endothelial cell end-feet) following myocyte contraction. 4) In the brain, the role of astrocytes in neurovascular coupling has recently been clarified. Until now, it was believed that activation of only BKCa channels in astrocytic end-feet (encircling the cerebral arterioles) generated K+ clouds in the astro-myocyte spaces. Subsequent activation of myocyte inward rectifiers generated hyperpolarisation and vessel dilatation. It is now clear that not only BKCa but also IKCa channels (both on astrocytic end-feet) are involved in this neurovascular coupling phenomenon.