The effects of hormone replacement therapy (HRT) on thrombosis risk, thrombotic variables, and the inflammatory marker C-reactive protein (CRP) may vary by route of administration (oral versus transdermal). We studied the relationships of 14 thrombotic variables (previously related to cardiovascular risk) and CRP to menopausal status and to use of HRT subtypes in a cross-sectional study of 975 women aged 40-59 years. Our study confirmed previously-reported associations between thrombotic variables and menopausal status. Oral HRT use was associated with increased plasma levels of Factor IX, activated protein C (APC) resistance, and CRP; and with decreased levels of tissue plasminogen activator (t-PA) antigen and plasminogen activator inhibitor (PAI) activity. Factor VII levels were higher in women taking unopposed oral oestrogen HRT. The foregoing associations were not observed in users of transdermal HRT; hence they may be consequences of the "first-pass" effect of oral oestrogens on hepatic protein synthesis. We conclude that different effects of oral and transdermal HRT on thrombotic and inflammatory variables may be relevant to their relative thrombotic risk; and suggest that this hypothesis should be tested in prospective, randomised studies.