We have measured the excretion of a major urinary metabolite of thromboxane B2 (TxB2), i.e., 2,3-dinor-TxB2, during the infusion of exogenous TxB2 over a 50-fold dose range to enable estimation of the rate entry of endogenous TxB2 into the bloodstream. Four healthy male volunteers received 6-h i.v. infusions of venhicle alone and TxB2 at 0.1, 1.0, and 5.0 ng/kg X min in random order. They were pretreated with aspirin at a dose of 325 mg/d in order to suppress endogenous TxB2 production. Urinary 2,3-dinor-TxB2 was measured before, during, and up to 24 h after the infusions and in aspirin-free periods, by means of radioimmunoassay. The nature of the extracted immunoreactivity was characterized by thin-layer chromatography and confirmed by negative ion-chemical ionization gas chromatography/mass spectrometry. Aspirin treatment suppressed urinary 2,3-dinor-TxB2 excretion by 80%. The fractional elimination of 2,3-dinor-TxB2 was independent of the rate of TxB2 infusion and averaged 5.3 +/- 0.8%. Interpolation of metabolite values obtained in aspirin-free periods onto the linear relationship between the quantities of infused TxB2 and the amount of metabolite excreted in excess of control values (y = 0.0066x, r = 0.975, P less than 0.001) permitted calculation of the mean rate of entry of endogenous TxB2 into the circulation as 0.11 ng/kg X min. The rate of disappearance of immunoreactive TxB2 from the circulation was monoexponential over the first 10 min with an apparent half-life of 7 min. This corresponded to a maximal estimate of the plasma concentration of endogenous TxB2 of 2.0 pg/ml. These results suggest that ex vivo platelet activation and/or analytical problems confound estimates of endogenous thromboxane release based on plasma TxB2 and provide a rationale for seeking longer-lived enzymatic metabolites of TxB2 in plasma.