Recent studies have suggested a differential influence of mean pressure and pulse pressure on myocardial infarction and stroke, and differences among the major drugs in their efficacy at preventing these individual endpoints. We hypothesized that antihypertensive drugs have differing influences upon the pulse wave even when their effects on blood pressure are the same. We studied 30 untreated hypertensive patients, aged 28-55 years, who were rotated through six 6-week periods of daily treatment with amlodipine 5 mg, doxazosin 4 mg, lisinopril 10 mg, bisoprolol 5 mg, bendrofluazide 2.5 mg or placebo. The best drug was repeated at the end of the rotation. Blood pressure readings and radial pulse tonometry (by Sphygmocor) were performed at each visit, and blood was taken for measurement of levels of atrial natriuretic peptide and brain natriuretic peptide (BNP). The Sphygmocor derivation of the central aortic pulse wave was used to measure time for transmission of the reflected wave (T(R)) and the augmentation index (AI), which is the proportional increase in systolic pressure due to the reflected wave. There was a dissociation between the effects of the drugs on blood pressure and pulse wave analysis. Bisoprolol caused the greatest falls in blood pressure and T(R), but was the only drug to increase AI. This paradoxical response to bisoprolol was associated with a 3-fold increase in plasma BNP levels. There was a smaller elevation of BNP in women compared with men, as described previously, and this elevation also was associated with significantly higher values of AI. Other drugs reduced AI, and this was associated with a significant decrease in BNP by amlodipine. In conclusion, antihypertensive drugs differ in their short-term effects on augmentation of the systolic pulse wave and secretion of BNP from the heart, regarded as a sensitive measure of strain on cardiomyocytes. These differences may help to explain cause-specific differences in outcome in recent trials.