Fifteen migraine patients were administered 2 mg ergotamine tartrate in a partial cross-over design as a single, oral tablet, rectal suppository and rectal solution. Eight of these patients were in a previous investigation given 0.5 mg ergotamine tartrate intravenously. The blood samples were taken up to 54 h after oral and suppository while it was followed for only 3 h after rectal solution. The chemical analysis was performed by applying h.p.l.c. method with a limit of sensitivity of 0.1 ng/ml ergotamine base in plasma. No ergotamine was detected in the blood samples after the oral route, whereas small and very variable quantities was found in blood after the rectal route. Regular calculation of bioavailability could therefore not be performed. An estimate of the maximal possible bioavailability was found to yield a mean value of 2% (tablets); 5% (suppositories) and 6% (rectal solution). Rectal solution elicited faster absorption and the extent of absorption was significantly higher (P less than 0.05) than for the suppository.