PURPOSE. The study aimed at assessing if the European guideline on the use of antihypertensive drugs (AD) in pregnancy are followed in clinical practice. We also evaluated the association between the use of non-recommended drugs and individual characteristics. METHODS. This study analyzed a cohort of 86 171 singleton deliveries occurring between 2009-2010 in the Lombardy region, Italy. Women with first prescription of AD during pregnancy were considered as incident users. Methyldopa, labetalol and nifedipine were considered as "recommended drugs"; all other AD were considered as "non-recommended". Odds Ratio and 95% confidence intervals were estimated. RESULTS. Among the 1009 patients (1.2%) exposed to AD during pregnancy, 675 (66.9%) were incident users. Among the incident users, 31% received non-recommended drugs; this proportion decreased to 18% among women who started treatment in the third trimester. Women with at least four concomitant diseases had an elevated risk of receiving non-recommended drugs in pregnancy (OR 2.68; 95% CI 1.10-6.73). CONCLUSIONS. Exposure to recommended antihypertensives increased during pregnancy. Nevertheless, a fraction of users that continued or began treatment with non-recommended medications was still present.