This study will investigate the non-Gaussian nature of speckle found in ultrasonic images through the use of phantoms and analytical models. Analyses of the envelope, phase and correlation of the echoes from the scatterers will be used to develop a general model for ultrasonic scattering. Using the model, tissue classification schemes will be suggested along with qualitative and quantitative evaluation of these schemes. A processor to suppress speckle also will be implemented. This research is important because the results could help improve tissue classification and, in particular, the ability to distinguish normal from abnormal tissue using ultrasound.