A previous regression analysis of flood peaks on urban basins in St. Louis County, Missouri, indicated that the basin characteristics of percentage of impervious area and drainage area were statistically significant for estimating the 2-, 5-, 10-, 25-, 50-. and 100-yr peak discharges at ungaged urban basins. In this statewide regression analysis of the urban basins for Missouri, an alternative basin characteristic called the percentage of developed area was evaluated. A regression analysis of the percentage of developed area (independent variable), resulted in a simple equation for computing percentage of impervious area. The percentage of developed area also was evaluated using flood-frequency data for 23 streamflow gaging stations, and the use of this variable was determined to be valid. Using nationwide data, an urban basin characteristic known as the basin development factor was determined to be valid for inclusion in urban regression equations for estimating flood flows. The basin development factor and the percentage of developed area were compared for use in regression equations to estimate peak flows of streams in Missouri. The equations with the basin development factor produced peak flow estimates with slightly smaller average standard errors of estimate than the equation with the percentage of developed area; however, this study indicates that there was not enough statistical or numerical difference to warrant using the basin development factor instead of the percentage of developed area in Missouri. The selection of a basin characteristic to describe the physical conditions of a drainage basin will depend not only on its contribution to accuracy of regression equations, but also on the ease of determining the characteristics; the percentage of developed area has this advantage. A correlation analysis was made by correlating drainage area to percentage of impervious area, the percentage of developed area, and the basin development factor. The results of the analysis indicate that the three basin characteristics are independent of drainage area and appropriate to use in multiple-regression analysis. (Author 's abstract)