New atmospheric particles with diameters of 3–10 nm and their subsequent growth to cloud condensation nucleus have been observed at various places in the European boundary layer. These events have been observed simultaneously within wide geographical areas (over 1000 km) in connection to specific weather systems, the cold air behind cyclones. Here we show that atmospheric aerosol formation (i.e. nucleation and initial growth) is favoured by the outbreak of cold Arctic air over northern Europe. Aerosol formation was about twice as common in Arctic air as in sub-Polar air, and even more so compared to other air masses. The most important general factor favouring aerosol formation in Arctic air and marine air was weaker competing condensational sink (CS) for the precursor gases (less pre-existing aerosols), while high CS prevented aerosol formation in heated sub-Polar air and mid-latitude air. High SO₂ levels favoured nucleation in continental air and high UV-B radiation in sub-tropical air. The critical factor that determined if aerosol formation would start on a day with Arctic air was the UV-B radiation. The same applied to sub-Polar air and continental air, while increased SO₂ concentration could trigger formation in heated sub-Polar and mid-latitude air, and reduced CS could cause formation in mid-latitude, marine or mixed/transient air. We speculate that strong emissions of volatile organic compounds from the Boreal forest and strong boundary layer dynamics may have caused aerosol formation in sub-Polar air masses and air in transition from a marine to a continental character. The monthly frequency of Arctic air masses and the probability for photo-chemically driven aerosol formation explains the observed annual cycle in monthly particle formation frequency as well as much of the inter annual variability. The same cyclones that transport cold, clean air from the Arctic to Europe will also transport warm polluted air in the other direction, which help cause the Arctic Haze phenomena. The cyclones have a key role for the atmospheric aerosol life cycle in mid to high latitudes. Due to the observed growth to the size of CCN in one to two days, there is a potential feed back from the effects on the CCN population and cloud albedo even within the same weather system, but also on the climatic time scale.