BACKGROUND: Higher blood pressure (BP) values in cold than in hot months has been documented in hypertensives. These changes may potentially contribute to the observed excess winter cardiovascular mortality. However, the association with weather has always been investigated by considering the relationship with a single variable rather than considering the combination of ground weather variables characterizing a specific weather pattern (air mass (AM)). METHODS: We retrospectively investigate in Florence (Italy) the relationship between BP and specific AMs in hypertensive subjects (n = 540) referred to our Hypertension Unit for 24-h ambulatory BP monitoring during the period of the year characterized by the highest weather variability (winter). Five different winter daily AMs were classified according to the combination of ground weather data (air temperature, cloud cover, relative humidity, atmospheric pressure, wind speed, and direction). RESULTS: Multiple variable analysis selected the AM as a significant predictor of mean 24-h BP (P < 0.01 for diastolic BP (DBP) and P < 0.05 for systolic BP (SBP)), daytime DBP (P < 0.001) and nighttime BP (P < 0.01 for both SBP and DBP), with higher BP values observed in cyclonic (unstable, cloudy, and mild weather) than in anticyclonic (settled, cloudless, and cold weather) days. When the association with 2-day sequences of AMs was considered, an increase in ambulatory BP followed a sudden day-to-day change of weather pattern going from anticyclonic to cyclonic days. CONCLUSIONS: The weather considered as a combination of different weather variables may affect BP. The forecast of a sudden change of AM could provide important information helpful for hypertensives during winter.