As a new approach, urban air pollution was characterised by the variation of columnar content of the pollutants. Columnar content (CC) was estimated as the product of the pollutant's mixing ratio and the mixing height. Mixing ratio data of the Metropolitan Air Quality Monitoring Network of Budapest were used, whilst mixing height was calculated by the meteorological AERMET model code. Time variation of CC refers to the real pollution exchange in the atmosphere that allows direct investigation of the emissions as well as post-emission modifications of the pollutants (such as chemical degradation or production). The diurnal urban CO cycle was found to be determined by two or three main influx peaks according to the traffic pattern of the site. The diurnal variation of NO x level was found to be driven by traffic emission as well. Variant ratios of NO x to CO influxes obtained for the different locations of the city range from 0.12 to 0.23, probably according to the vehicle composition of the traffic. The daily balance of photochemical production, chemical degradation and deposition of ozone yielded negative or positive depending on the location. Negative balances imply that the polluted urban atmosphere is a net ozone neutraliser source. Entrainment from the free troposphere yielded the major contributor to the diurnal ozone level at each site. The diurnal urban PM 10 cycle was found to be determined by traffic emission during the morning and evening rush hours whilst secondary aerosol formation around noon. In the evening, high PM 10 level rise was observed due to direct traffic emission as well as rapid conversion of the fine aerosol fraction to the coarse fraction.