Mobility of population is one of major problems in large metropolitan areas. It affects different social, economic and environmental aspects of a city’s life. Urban mobility is characterised by highly dynamic spatial-temporal variability not detected by census or surveys. Mobile phones allow tracking of the population, giving information about their time–space location. A multi-city population mobility study was carried out in seven main Italian large metropolitan areas for 2 months (1st March to 30th April, 2015), starting from high-resolution gridded population data derived from mobile phone traffic data. Mean daily–nightly variations of population density were calculated during weekdays with increments of population of about 25% on average and peaks up by 200% in hotspot areas. Strong spatial gradients were identified with both accumulation and depletion of population in urban zones of geographical dimensions of 55 and 44%, respectively, of the municipality areas on average. Larger metropolitan areas, such as Rome, Milan, and Turin, show the highest increments of population, while medium-sized cities (e.g., Bari and Palermo) exhibit lower values and weaker gradients of variation of population. By means of a developed mobility factor, three main mobility phenomena (morning and evening commuting, diffuse mobility) occurring during weekdays were identified and characterized.