This paper compares the performance of various statistical downscaling methods for daily precipitation data in six regions across the United States. The methods compared include different weather generator techniques (WGEN and SPEL), two methods using grid point vorticity data (B-Circ and C-Circ), and two artificial neural network (ANN) approaches. The downscaling models were calibrated using observed and general circulation model (GCM) generated daily precipitation time series. The GCM used was the U.K. Meteorological Office's Hadley Centre's coupled ocean-atmosphere model (HadCM2) forced by CO2 and sulfate aerosol changes. The study found significant differences in the skill of the downscaling methods, with the weather generation techniques yielding the smallest differences between observed and simulated daily precipitation. The ANN methods performed poorly due to their inability to simulate wet-day occurrence statistics adequately. Changes in precipitation between the present (1980-1999) and future (2080-2099) scenarios produced by the statistical downscaling methods were generally smaller than those produced directly by the GCM. The results suggest that additional predictor variables beyond vorticity are needed to confidently downscale future climate scenarios. The study also highlights the need for spatial autocorrelation considerations, improvements in low-frequency variability simulations, and the inclusion of more predictor variables to enhance the robustness of downscaling models.This paper compares the performance of various statistical downscaling methods for daily precipitation data in six regions across the United States. The methods compared include different weather generator techniques (WGEN and SPEL), two methods using grid point vorticity data (B-Circ and C-Circ), and two artificial neural network (ANN) approaches. The downscaling models were calibrated using observed and general circulation model (GCM) generated daily precipitation time series. The GCM used was the U.K. Meteorological Office's Hadley Centre's coupled ocean-atmosphere model (HadCM2) forced by CO2 and sulfate aerosol changes. The study found significant differences in the skill of the downscaling methods, with the weather generation techniques yielding the smallest differences between observed and simulated daily precipitation. The ANN methods performed poorly due to their inability to simulate wet-day occurrence statistics adequately. Changes in precipitation between the present (1980-1999) and future (2080-2099) scenarios produced by the statistical downscaling methods were generally smaller than those produced directly by the GCM. The results suggest that additional predictor variables beyond vorticity are needed to confidently downscale future climate scenarios. The study also highlights the need for spatial autocorrelation considerations, improvements in low-frequency variability simulations, and the inclusion of more predictor variables to enhance the robustness of downscaling models.