In a medline search (covering 1966 to Sept. 1996) 32 clinical studies were identified, in which the efficacy of paracetamol or matamizol per se, or in comparison to other analgesics, in various chronic pain states, such as migraine, dysmenorrhoea, arthritis and osteoarthritis pain and cancer pain had been examined. In patients with migraine (4 studies) several other analgesics (ibuprofen, mefenamic acid, flupirtin) were slightly more effective than paracetamol, however, the efficacy of paracetamol itself had not been assessed. In patients with chronic tension headache (1 study) paracetamol was superior to placebo, but less effective than naproxen. Pain of dysmenorrhoea was not, or only marginally improved by paracetamol in 3 studies, efficacy was reported in 1 study. Similarly, pain in rheumatoid arthritis was not significantly alleviated when paracetamol was given alone (3 studies) and marginally improved, when combined with naproxen and tested against naproxen alone (1 study). Some improvement by paracetamol of pain scores in patients with osteoarthritis (5 studies) requires further clinical confirmation. No studies were found, in which metamizol had been studied in chronic non-cancer pain. Paracetamol and/or metamizol have been included in 14 studies on cancer pain, most of these studies attempting to validate the WHO analgesic ladder for cancer pain treatment. However, except for one study, in which metamizol was comparable in efficacy to morphine, all other publications do not provide detailed information on the efficacy of individual analgesics. Therefore it is not possible at present, to assess the possible merits of paracetamol or metamizol in the treatment of cancer pain from published studies.