Predictions of global circulation models (GCMs) that account for increasing concentrations of greenhouse gases and aerosols in the atmosphere show that warming in the Arctic will be amplified in response to the melting of sea ice and snow cover. There is now conclusive evidence that much of the Arctic has warmed in recent decades. Northern Alaska is one region where significant warming has occurred, especially during winter and spring. We investigate how the changing climate of northern Alaska has influenced the annual cycle of snow cover there and in turn, how changes in snow cover perturb the region's surface radiation budget and temperature regime. The focus is on Barrow, Alaska, for which comprehensive data sets exist. A review of earlier studies that documented a trend toward an earlier disappearance of snow in spring is given. Detection and monitoring activities at Barrow are described, and records of snow disappearance from other sites in the Alaskan Arctic are compared. Correlated variations and trends in the date of final snowmelt (melt date) are found by examining several independent time series. Since the mid-1960s the melt date in northern Alaska has advanced by ~;8 days. The advance appears to be a consequence of decreased snowfall in winter, followed by warmer spring conditions. These changes in snowfall and temperature are attributed to variations in regional circulation patterns. In recent decades, there has been a higher frequency of northerly airflow during winter that tends to diminish snowfall over northern Alaska. During spring, however, intrusions of warm moist air from the North Pacific have become more common, and these tend to accelerate the ablation of snow on the North Slope of Alaska. One result of an earlier melt date is an increase in the net surface radiation budget. At Barrow, net radiative forcing can exceed 150 W m-2 on a daily basis immediately following the last day of snowmelt, and as a result of an 8-day advance in this event, we estimate an increase of ~2 W m-2 on an annual basis. Our results are in general agreement with earlier analyses suggesting that reductions in snow cover over a large portion of the Arctic on an annual basis have contributed to a warming of the Northern Hemisphere (NH). In addition, the terrestrial ecosystems of the region are very sensitive to snow cover variations. There is growing concern that these perturbations are anthropogenically forced and adapting to these environmental changes will have significant social and economic consequences. While observed decreases in NH snow cover are in broad agreement with GCM simulations, our analyses suggest that internal (or natural) shifts in circulation patterns underlie the observed variations. Continued monitoring and further study is needed to determine whether the earlier disappearance of snow cover in spring in northern Alaska is an indicator of greenhouse-forced global warming or is a manifestation of a more natural, long-term cycle of climate change.