A generalized skew coefficient analysis was completed for South Dakota to test the validity of using the generalized skew coefficient map in Bulletin 17B of the 1982 United States Water Resources Council, ?Guidelines for Determining Flood Flow Frequency.? Results of the analysis indicate that the Bulletin 17B generalized skew coefficient map generally provides adequate generalized skew coefficients for estimating peakflow magnitudes and frequencies for South Dakota gaging stations. Peak-flow records through 1994 for 197 continuous- and partial-record streamflow-gaging stations that had 10 or more years of unregulated systematic record were used in a generalized leastsquares regression analysis that relates peak flows for selected recurrence intervals to selected basin characteristics. Peak-flow equations were developed for recurrence intervals of 2, 5, 10, 25, 50, 100, and 500 years for seven hydrologic subregions in South Dakota. The peak-flow equations are applicable to natural-flow streams that have drainage areas less than or equal to 1,000 square miles. The standard error of estimate for the seven hydrologic subregions ranges from 22 to 110 percent for the 100-year peak-flow equations. Weighted peak flows for various frequencies based on gaging-station data and the regional regression equations are provided for each gaging station. Examples are given for (1) determining peak-flow magnitudes and frequencies for ungaged sites on ungaged streams; (2) determining weighted peak-flow magnitudes and frequencies for gaging stations; and (3) using the drainage-area ratio method for determining peakflow magnitudes and frequencies for ungaged sites near a gaging station on the same stream and ungaged sites between two gaging stations on the same stream.