A method based on broad-band amplitude for obtaining attenuation and echogenicity estimates from homogeneous phantoms and tissues is described. Although a number of assumptions about the beam and scattering properties must be made, the attenuation results are as accurate as those obtained by the spectral-difference method and show less variability. The method was applied to the livers of 18 healthy volunteers and 76 patients with liver disease, including 29 patients with chronic hepatitis and 30 patients with Gaucher disease. Patients with chronic hepatitis formed a bimodal distribution, with one group having lower than normal attenuation and echogenicity values and another having elevated values. Six patients with type I glycogen storage disease and fatty infiltration of the liver had highly attenuating, echogenic livers as did patients with fatty livers from other causes. This finding confirms the subjective impressions already reported in the literature. In contrast to previous literature reports, patients with cirrhosis had normal attenuation and echogenicity values. Reasons for this discrepancy are discussed.