The aim of this study was to determine the impact of the administration route and cigarette smoking on plasma oestrogen levels during oral and parenteral oestrogen replacement therapy (ERT). Fourteen healthy postmenopausal women (six smokers and eight non-smokers) were recruited for a prospective, randomised, crossover study at a private outpatient medical centre in Oslo, Norway. All patients were randomised to receive cyclic therapy with oestradiol and norethisterone orally or by the transdermal route each for a 6-month period. Plasma levels of oestrone (Oe 1 ), oestradiol (Oe 2 ) and oestrone sulphate (Oe 1 S) were determined using highly sensitive RIA methods before and during hormone replacement therapy given by the oral and transdermal route. Comparing smokers and nonsmokers, plasma levels of Oe 1 , Oe 2 and Oe 1 S were all found to be 40-70% lower in smokers compared with non-smokers when ERT was given orally (Oe 1 S, P<0·05; Oe 1 and Oe 2 , P<0·01 for both). Oe 2 given orally caused a higher Oe 1 S/Oe 2 ratio but also a higher Oe 1 /Oe 2 ratio compared with parenteral therapy in smokers (40·2 versus 7 . 0, P<0·01; and 3·2 versus 0·8, P<0·05 respectively). No significant differences in these parameters in the different test-situations were seen in non-smokers. Except for a lower level of Oe 1 S in smokers (non-significant), no difference in plasma oestrogen levels between smokers and non-smokers was observed during parenteral therapy. In conclusion, cigarette smoking has been shown to have major impact on plasma oestrogen levels during oral but not during parenteral Oe 2 replacement.