Aerosol data consisting of condensation nuclei (CN) counts, black carbon (BC) mass concentration, and aerosol light scattering coefficient at the wavelength of 450 nm (SC) measured at Barrow, AK, from 1986 to 1997 have been analyzed. BC and SC show an annual cycle with the Arctic haze maxima in the winter and spring and the minima in the summer. The CN time series shows two maxima in March and August. Potential source contribution function (PSCF) that combines the aerosol data with air parcel backward trajectories was applied to identify potential source areas and the preferred pathways that give rise to the observed high aerosol concentrations at Barrow. Ten-day isentropic back trajectories arriving twice daily at 500 and 1500 m above sea level were calculated for the period from 1986 to 1997. The PSCF analyses were performed based on the 80th percentile criterion values for the 2- and 24-h averages of the measured aerosol parameters. There was a good correspondence between PSCF maps for the 2- and 24-h averages, indicating that 1-day aerosol sampling in the Arctic adequately represents the aerosol source areas. In winter, the high PSCF values for BC and SC are related to industrial source areas in Eurasia. The trajectory domain in winter and spring is larger than in summer, reflecting weaker transport in summer. No high PSCF areas for BC and SC can be observed in summer. The result is related to the poor transport into the Arctic plus the strong removal of aerosol by precipitation in summer. In contrast to the BC and SC maps, the CN plot for summer shows high PSCF areas in the North Pacific Ocean. High CN values appearto be mostly connected with the long-range transport from Eurasia in winter and spring and with the reduced sulfur compound emission from biogenic activities in the ocean in the summer. PSCF analysis was found to be effective in identifying potential aerosol source areas.