We have investigated the effects of a low-dose aspirin regimen (120 mg orally, then 20 mg twice daily) on the in vivo synthesis of prostacyclin, thromboxane and prostaglandin E in man by measurement of their urinary metabolites (PGI 2-M, TxB 2-M, PGE-M) using gas chromatography-mass spectrometry. A comparison was made between the selectivity of low-dose aspirin for thromboxane vs prostacyclin synthesis in patients with atherosclerosis, age-matched controls without vascular disease, and young healthy volunteers. After one week of treatment, aspirin reduced TxB 2-M synthesis to a similar extent in the three groups (mean declines of 86, 84 and 78% respectively), while there was an unexpected difference in effect on PGI 2-M, with only a 27% fall in the young volunteers but 53% and 54% declines in the patients with vascular disease and their age-matched controls. Serum TxB 2 was reduced >98% in all groups while PGE-M excretion was unchanged. These results indicate that bioselectivity for inhibition of Tx synthesis by aspirin is more difficult to achieve in older subjects than in the young volunteers previously studied and that very low, frequent dosing, or a sustained-release preparation of aspirin would be needed to achieve bioselectivity for Tx inhibition in patients with vascular disease.