Enzymatic metabolites of arachidonic acid (eicosanoids) have potent biologic actions in vitro that suggest their pathophysiologic importance in vivo. To address this possibility, analytic methodology has been developed to permit study of the formation of these compounds in vivo. Both radioimmunoassay and gas chromatography-mass spectrometry have been used to measure stable but biologically inactive metabolites of the eicosanoids. Although indirect, such measures are presently the most reliable, because superfusion-bioassay lacks the specificity and precision necessary for quantitative analysis of eicosanoid formation in vivo. Measurement of eicosanoids and their hydration products and metabolites in urine represents a non-invasive approach to the assessment of eicosanoid biosynthesis. Although a tissue of origin cannot be ascribed definitely to a compound measured in urine, corroborative evidence can be obtained to indicate the predominant tissue source under physiologic and pathologic conditions. This relates particularly to the distinction between renal and extrarenal biosynthesis of these compounds. Although similar limitations apply to the measurement of eicosanoids in plasma, these may also be confounded by sources of artifact related to blood withdrawal. In the case of thromboxane B2, these concerns have been addressed by the development of methods to measure its enzymatic metabolites in plasma. Finally, formation of eicosanoids may be studied in localized compartments such as lavage or synovial fluid. Such an approach has recently provided biochemical evidence for increased formation of prostacyclin and prostaglandin E2 at the platelet-vascular interface during selective inhibition of thromboxane synthase in humans.