Cardiac fibrosis and high levels of circulating collagen markers has been associated with left ventricular (LV) hypertrophy. However, the relationship to vascular hypertrophy and blood pressure (BP) load is unclear. In 204 patients with essential hypertension and electrocardiographic LV hypertrophy, we measured sitting BP, serum collagen type I carboxy-terminal telopeptide (ICTP) reflecting degradation, procollagen type I carboxy-terminal propeptide (PICP) reflecting synthesis and LV mass by echocardiography after 2 weeks of placebo treatment and after 1 year of antihypertensive treatment with a losartan-or an atenolol-based regimen. Furthermore, we measured intima-media thickness of the common carotid arteries (IMT), minimal forearm vascular resistance (MFVR) by plethysmography and ambulatory 24-h BP in around half of the patients. At baseline, PICP/ICTP was positively related to IMT (r ¼ 0.24, Po0.05), MFVR men (r ¼ 0.35, Po0.01), 24-h systolic BP (r ¼ 0.24, Po0.05) and 24-h diastolic BP (r ¼ 0.22, Po0.05), but not to LV mass. After 1 year of treatment with reduction in systolic BP (175715 vs 151717 mmHg, Po0.001) and diastolic BP (9978 vs 8879 mmHg, Po0.001), ICTP was unchanged (3.771.4 vs 3.871.4 lg/l, NS) while PICP (121739 vs 102729 lg/l, Po0.001) decreased. The reduction in PICP/ICTP was related to the reduction in sitting diastolic BP (r ¼ 0.31, Po0.01) and regression of IMT (r ¼ 0.37, Po0.05) in patients receiving atenolol and to reduction in heart rate in patients receiving losartan (r ¼ 0.30, Po0.01). In conclusion, collagen markers reflecting net synthesis of type I collagen were positively related to vascular hypertrophy and BP load, suggesting that collagen synthesis in the vascular wall is increased in relation to high haemodynamic load in a reversible manner.