Environmental noise data, including 24-hour noise surveys, taken at various urban locations throughout the United States have been compiled and compared. The data were taken from the open literature and also from the files of the U.S. Department of Housing and Urban Development. It is suggested that the urban noise scene is sufficiently different from the nonurban situation so that, prediction and evaluation schemes commonly used in environmental impact statements are impractical or invalid. For example, relatively sophisticated approaches to traffic noise impact, e.g., the Highway Research Board's design guide (Rep. 117), prove to be intractable for moderately sized projects in urban areas. However, an analysis of the data infers alternate approaches and these are discussed.