Abstract: The paper presents a method of lake bottom sediment classification using echo signals provided by a wideband active sonar.The five types of sediments are silt.rocks.pebbles.sand.and mixture of sand and gravel.Different sub-band divisions for frequency domain feature extraction are compared and it is shown that the contant Q method provides better results in comparison with the constant bandwidth method.Using a standard feedforward BP network,82.1% correct classification in average has been achieved for the five classes of sediments.The advantages of using wideband sonar signals to improve the classification Performance are also shown.