To assess the impact of sumatriptan in clinical practice, we undertook a retrospective analysis of the government of Newfoundland and Labrador's prescription drug program data base for 35 consecutive patients prescribed sumatriptan. The number of doses of all drugs prescribed ranged from 121 to 18874 on from 4 to 357 prescriptions per patient over 1 to 19 months. The mean number of doses of analgesic drugs prescribed before sumatriptan therapy was 56 per month and after initiation of sumatriptan was 46 per month. The prescribing of multiple analgesics was common; 79% received three or more different analgesics. Twenty-two (63%) patients were prescribed medications indicated for the prophylaxis of migraine concomitantly with drugs indicated for symptomatic treatment. Twenty-four (69%) patients were prescribed medication capable of inducing migraine. We conclude that sumatriptan did not have a major impact on the outcomes of these patients judged by their use of analgesics. The simplest explanation is that many of the patients were suffering from analgesic-induced headache rather than migraine. In addition, we conclude that there were deficiencies in prescribing practices including numbers, quantities, and choice of analgesics; the use of analgesics concomitantly with drugs indicated for migraine prophylaxis; and the use of drugs capable of inducing migraine. Further research is required to validate these findings.