In order to identify suitable parameters for measurement of thromboxane production in vivo, the metabolism of TXB2 was studied in the human. [3H8]-TXB2 was given intravenously to a healthy human volunteer. Blood samples were collected for 50 min after the injection, and urine was collected for 24 hours. The urinary and blood metabolic profiles were visualized by the use of two-dimensional TLC and autoradiography. Identification of metabolites was achieved with GC/MS and in some cases by cochromatography with reference compounds in TLC and GC. In blood, unmetabolized TXB2 was the dominating compound during the first 30 min. Three less polar metabolites appeared, two of which were identified as 11-dehydro-TXB2 and 11,15-didehydro-13,14-dihydro-TXB2, respectively. The third compound was tentatively identified as 15-dehydro-13,14-dihydro-TXB2. Since 11-dehydro-TXB2 was one of the major metabolites in blood as well as urine, it was deemed suitable as target for measurement of thromboxane production in vivo. The advantages of 11-dehydro-TXB2 over its parent compound, TXB2, were demonstrated in experiments where unlabeled TXB2 was injected i.v. to a human volunteer, and the blood and urinary levels of both compounds were then followed by radioimmunoassay. Measured levels of 11-dehydro-TXB2 were found to give a more reliable picture of metabolic events than TXB2, the latter compound to a large extent reflecting technical difficulties during blood sample collection.