OBJECTIVE: To quantify and qualify the antihypertensive pharmaceutical expenditure incurred for the treatment of a population in the real world. METHODS: An administrative database, collecting information on all purchased drugs, was used to perform a retrospective analysis for each health-assisted subject by Ravenna Local Health Unit. By crosschecking with the registry office and hospital database, gender, age and previous hospitalizations for cardiovascular diseases were also made available. All new users over 20 years receiving a first prescription for diuretic, beta-blocker, calcium-channel-blocker, angiotensin-converting-enzyme inhibitor or angiotensin II antagonist between 01/01/1997 and 12/31/1997 were included. According to persistency of treatment, patients were classified as continuers (duration of therapy over 273 days with the initial antihypertensive class), switchers (duration of therapy over 273 days with a different class of antihypertensive than originally prescribed) and discontinuers (duration of therapy less than 273 days). The follow-up period lasted 365 days. RESULTS: A total of 16,783 patients were enrolled, including 7,409 men (44.1%) and 9,374 women (55.9%) with an average age of 56.1±18.3 years. Continuers and switchers represented 26.9% and 8.2% of enrolled subjects, respectively. The percentage of treatment interruptions (64.9%) ranged from 42.0% with angiotensin II antagonists to 69.8% with diuretics, and53.8% of the population interrupted therapy after the first prescription. Age, co-morbidities and previous hospitalizations were higher for continuers and switchers than for discontinuers. The overall expenditure accounted for € 1,076,053.55 of which 54.8% was for continuers, 19.8% for switchers and 25.4% for discontinuers. The percentage of the overall cost for discontinuers ranged from 14.8% with angiotensin II antagonists to 30.6% with calcium-channel blockers. CONCLUSION: The high percentage of discontinuers translated into an expense incurred improperly. The high percentage of interruptions after the first prescription seemed to indicate that medication prescribed at enrollment was not appropriate.