Using a tissue-mimicking phantom and a computerised sonography system an ultrasound b scan analysis was performed. The images were acquired with different fat-muscle combinations as ultrasound wave path. These are simulations of a different composition or thickness of the abdominal wall. The analysis of the ultrasound images was performed using mathematical statistical parameters for ultrasound tissue characterisation. We examined the influence of the thickness or composition of the abdominal wall on ultrasound image brightness, microtexture and macrotexture. A dependence of ultrasound image brightness and microtexture on the thickness of the fat path was observed. Muscle tissue shows only a weak influence on the microtexture of the ultrasound image. These data demonstrate that a thick fat layer leads to a darker and diffuse image. This disturbs the evaluation e.g. of the liver parenchyma by analysis of the echo structure and may lead to false diagnoses especially in adiposity.