On the basis of a 1-year double-blind comparative study of doxazosin and atenolol in hypertensive patients, we have previously described sustained and significantly different drug effects on the blood lipid profile. The present paper presents a retrospective analysis of the smoking subgroups for each drug regimen, for both the original 12-month double-blind period and for a further 12-month open-label period. The changes in blood lipids were qualitatively and quantitatively similar in both the smoking subgroup and in the full study population (smokers and non-smokers). In the hypertensive smokers doxazosin increased high-density lipoprotein (HDL) cholesterol (+9.4%) and decreased triglycerides (-10.9%) whereas atenolol had the opposite effect, decreasing HDL cholesterol (-5.6%) and increasing triglycerides (+20.7%). The antihypertensive effects of doxazosin (n = 24) and atenolol (n = 23) were comparable over 24 months of treatment. Eighteen (75%) doxazosin and 17 (73.9%) atenolol patients reached the blood pressure goal (sitting diastolic blood pressure less than or equal to 90 mmHg with at least a 5-mmHg reduction, or a greater than or equal to 10-mmHg reduction from baseline). As a consequence of the lowered blood pressure and favourable lipid changes, doxazosin decreased the risk of coronary heart disease (as calculated from the Framingham risk equation) by -24.4% (P less than 0.05). However, the calculated risk was not significantly reduced by atenolol (-2.1%). It is concluded that the smoking status of the patient has no obvious implications for the safety, efficacy or metabolic effects of doxazosin.