594 blood donors were studied to define the influence of regular blood donation on blood pressure. Blood pressures were compared at the first phlebotomy performed at the Padova Blood Unit, the 13-year phlebotomy, and the median between the two. As a control population, a cohort of 594 non-donors from the same general population similar to the blood donors was studied twice at a 15-year interval. Subjects had made on average 7.3 +/- 0.4 donations at the first step, 23 +/- 0.6 at the second, and 30.9 +/- 0.7 at the last step. Unadjusted systolic blood pressure regularly and significantly increased by 1.6% from the first phlebotomy to the second one and by 2.3% from the first to third. Adjusted systolic blood pressure, on the contrary, significantly decreased by 2.1% and by 2.7% respectively. Diastolic blood pressure tended to increase insignificantly, while the adjusted one did not change at all. The increase of systolic blood pressure from the first to second step directly correlated with the number of donations, and this also applied to differences between the first and the third. When the increase in age was introduced in the multiple regression analysis, systolic blood pressure rise from the first to third step showed an inverse correlation with the number of phlebotomies. Our data render suspect the results of epidemiological investigations which took into consideration cohorts of blood donors; although these cohorts may be anagraphically representative of a general population, repeated phlebotomies introduce a bias leading to the detection of misleadingly low blood pressure values in regular blood donors.