In this paper we describe a recently developed technique to collect samples for atmospheric 222Rn measurements suited for deployment on (research) aircraft. Following a technical description of the sampler we present measurements from nine flights during the 1993 North Atlantic Regional Experiment (NARE) intensive over the North Atlantic Ocean and over the continent in the vicinity of Nova Scotia and New Brunswick, Canada. Samples were taken both in the boundary layer and in the free troposphere up to about 5500 m (msl). Concentrations ranged from 0.01 to 1.66 Bq/m3 in the free troposphere, with means of 1.66 and 0.52 Bq/m3, respectively. Boundary layer 222Rn concentrations were strongly correlated with back trajectory origins confirming the use of 222Rn as tracer for continental air masses. Samples collected in the marine boundary layer (MBL) were found to be significantly correlated with aerosol particle, NOy, and O3 concentrations reaffirming the utility of radon as a tracer of continental boundary layer air. Radon concentrations for the MBL samples were not found to be significantly correlated with transit times from the continent as estimated from back trajectory analysis implying that variations in the radon concentrations are dominated by factors other than radioactive decay. It is suggested that this factor is the height of the continental boundary layer into which the radon is emitted. On the basis of the radioactive decay we estimated apparent 222Rn ages for our samples and discuss the inherent difficulties in reliable age estimates. A comparison with hydrocarbon ratio age estimates for a highly polluted plume of air encountered during the campaign shows a similar age range for these samples and suggests that simultaneous measurements of species with different chemical and physical properties might allow narrowing the uncertainties in such ages estimates. ¿ American Geophysical Union 1996