The problem of discriminating between underwater mines and similarly sized false targets at ranges of several hundred meters remains a challenge. The Applied Physics Laboratory, University of Washington, in conjunction with Arete Engineering and Technologies Corporation, San Diego, have been developing detection and classification signal processing schemes using data collected in several shallow-water experiments with a broadband (2--20 kHz) sonar. The use of relatively low frequencies for mine hunting permits greater penetration into bottom sediments for improved detection of buried or partially buried mines. Wide bandwidth signals mitigate the lack of spatial resolution expected from a narrow-band analysis at such frequencies, and provide a wealth of information for classification. A brief description of the sonar and of the experiments conducted in Puget Sound using mines and minelike false targets will be given. Target responses as a function of time, frequency, and aspect angle will be presented, as will results of using several feature extraction methods to reduce the dimensionality of the data.