Significant concentrations of volatile organic compounds can be present in the interior of vehicles due to emissions from materials being part of the interior fittings and to combustion and evaporation emissions of the fuel. Two times in 1995 and 1996 carbon monoxide (CO), carbon dioxide (CO2), aromatic hydrocarbons, polycyclic aromatic hydrocarbons (PAH), diesel motor emission (estimated as elementary carbon) were determined in the inside of a car (a two year old VW - Golf with a three-way catalyst) and in a subway-train. On each sampling day (in total 16 daily measurements in the car and 16 in the subway) the substances were determined in the breathing zone of the passengers from 7 a.m. to 4 p.m. under different meteorologic conditions (winter- and summertime). The car route followed the subway from the western borough of Spandau to the south-east borough of Neukolln, there and back This sampling represented a realistic exposure model for driving in a high traffic and polluted area. The electric subway-train (also two years in use) connected the same parts of Berlin ( 31 km underground). At present only the results of the aromatic hydrocarbons, PAH (partial) and carbon monoxide can be presented in this paper. The mean values obtained during the two measurement periods inside the car were 22/22 mu g/m(3) (maximum 26/35 mu g/m(3)) for benzene, 69/90 mu g/m(3) for toluene, 38/69 mu g/m(3) for m-,p-xylene, 15/25 mu g/m(3) for o-xylene, 14/23 mu g/m(3) for ethylbenzene and 1.0/3.2 ng/m(3) for benzo(a)pyrene. The mean values obtained in the subway were 5.4/7.4 mu g/m(3) (maximum 7/16 mu g/m(3)) for benzene, 33/31 mu g/m(3) for toluene, 15/12 mu g/m(3) for m-,p-xylene, 5.4/4.7 mu g/m(3) for o-xylene, 5.5/4.4 mu g/m(3) for ethylbenzene and 0.7/4.0 ng/m(3) for benzo(a)pyrene. A comparison between subway and car shows significantly lower concentrations of the aromatic hydrocarbons and higher concentrations of PAH in the subway train. In wintertime the concentrations of benzo(a)pyrene are three to four times higher than in summer and followed the changing of the ambient air concentrations. Our results-in agreement with other studies-demonstrate that car occupants rather than subway passengers were exposed to motor vehicle exhaust and evaporative emissions. An overall risk assessment for benzene shows that - even under the conditions of only a one-hour car ride per day - this exposure represents nearly 30% of the total benzene risk. The mean levels of CO in the car were 6.1 +/- 3.9 ppm (summer) and 6.9 +/- 7.4 ppm (winter) respectively, with peak concentrations of 70 ppm. In the subway the values were 2.1 +/- 0.4 ppm (summer) and 1.9 +/- 1.9 ppm (winter). In the car it cannot totally be excluded that critical COHb-levels (2.5%, for patients with coronary heart disease) will be exceeded.