Understanding of the complex pharmacology of the spinal cord may lead to rational advances in pain treatment. It appears that a number of specific neurochemical mechanisms exist, by which spinally administered receptor selective agents may modify nociceptive transmission. Spinal administration of pure competitive N-methyl-D-aspartate (NMDA) antagonists affects only hyperpathic pain components, i.e. with signs of central sensitization, and most probably has a very limited role in postoperative pain treatment. On the other hand, it is well established that the non-competitive NMDA-antagonist ketamine gives good postoperative analgesia, probably by cerebral mechanisms also affecting other sensory modalities. Pure adenosine A1-receptor agonism at the spinal level mainly affects sensory allodynia to vibration, and is probably no alternative for postoperative pain treatment. In contrast, i.v. infusions of the non-selective A1/A2-receptor agonist adenosine given during a surgical procedure seem to decrease postoperative pain and requirements for postoperative analgesia. This apparent contradiction must be analysed further. Several drugs commonly used to treat postoperative pain, such as opioids, NSAIDs, ketamine and paracetamol, are linked to nitric oxide (NO) in their mechanism of action. The biosynthesis of NO in the central nervous system (CNS) is tonically involved in the nociceptive processing.