Recent studies using agonist analogues of ATP and other nucleotides have generated some surprising observations which may have ramifications for the classification of P2 receptors, particularly for those responses currently attributed to P2Y receptor activation. 2-MethylthioATP (2-MeSATP), the conventional P2Y receptor agonist, does not interact with ATP in the expected fashion in various models of endothelial function, suggesting that it acts by a different mechanism. Furthermore, in certain cell types where responses to ATP are mediated by phospholipase C activation, 2-MeSATP has little or no activity. Interestingly, the pyrimidine uridine triphosphate (UTP) invariably shows similar potency to ATP in systems where 2-MeSATP is inactive. In this article Steve O'Connor and colleagues discuss these data and their significance, and propose that separate receptors may be responsible: one sensitive to 2-MeSATP and the other, a 'nucleotide' receptor, sensitive to UTP.