In 1722 hypertensives recruited for the European Lacidipine Study on Atherosclerosis and treated with atenolol or lacidipine (±additional drugs) during 4 years, we evaluated the long-term reproducibility of dipping, nondipping, extreme dipping, and reverse dipping, an information of key relevance for defining the prognostic impact of these blood pressure (BP) phenotypes. Ambulatory BP was measured at baseline and every year during treatment, allowing repeated classifications of these conditions. Based on systolic BP, at baseline 50.1%, 37.5 %, 4.5 %, and 7.9 % were dippers, nondippers, extreme dippers, and reverse dippers, respectively. Antihypertensive drug treatment reduced the number of dippers and extreme dippers in favor of nondippers and reverse dippers. During treatment, the dipping and nondipping phenotypes were markedly unstable and shifts from one to other phenotypes were even more common in reverse and extreme dippers In only a very small fraction of patients, a given nighttime BP phenotype was persistently found throughout the 4 years. Data were similar for diastolic BP and in subgroups differing for the type of treatment, the treatment complexity, and the achieved on-treatment BP value Poor reproducibility characterized also the absolute nighttime BP reduction as measured by the nighttime BP fall and the night/day BP ratio. Thus nighttime BP phenotypes are all largely inconsistent, their detection over the entire treatment period involving just a minute fraction of the entire patients’ cohort.