OBJECTIVE: Although it is known that increased visit-to-visit or home day-by-day variability of blood pressure (BP), independently of its average value, results in an increased risk of cardiovascular events, the prognostic value of in-hospital day-by-day BP variability in secondary cardiovascular prevention has not yet been established. METHODS: We studied 1440 consecutive cardiac patients during a cardiovascular rehabilitation program of about 12 days after coronary artery bypass graft (CABG) and/or valve surgery. We measured auscultatory BP at the patient bed in each rehabilitation day twice, in the morning and the afternoon. We correlated SBP variability assessed as standard deviation (SBP-SD) and coefficient of variation (SBP-CoV) of the daily measures with overall mortality, cardiovascular mortality and major adverse cardiocerebrovascular events (MACCEs) after a mean follow-up of 49 months by Cox hazard analysis. RESULTS: In our patients (age 68 ± 11years, 61% hypertensive patients) the ranges of SBP-SD tertiles were: 4.1-9.1, 9.2-11.5 and 11.6-24.5 mmHg. Fifty-five percent of the patients underwent CABG, 33% underwent valve surgery, 12% both CABG and valve surgery. In CABG patients, the highest SBP-SD tertile showed the highest overall mortality, cardiovascular mortality and MACCEs (P < 0.01). Results remained significant after multivariate analysis adjusting for age, sex, mean SBP, BMI, hypertension, hyperlipidaemia, and diabetes. No association between SBP-SD and mortality or MACCEs was found in valve surgery patients. CONCLUSION: In-hospital day-by-day SBP variability predicts mortality and MACCEs in CABG patients, possibly representing a target during rehabilitation and treatment in secondary cardiovascular prevention.