Background: The echogenicity patterns of ultrasound scans contain information of tissue composition in muscles. The aim was: ( 1) to develop a quantitative ultrasound image analysis to characterize tissue composition in terms of intensity and structure of the ultrasound images, and ( 2) to use the method for characterization of ultrasound images of the supraspinatus muscle, and the vastus lateralis muscle. Methods: Computerized texture analyses employing first-order and higher-order grey-scale statistics were developed to objectively characterize ultrasound images of m. supraspinatus and m. vastus lateralis from 9 healthy participants. Results: The mean grey-scale intensity was higher in the vastus lateralis muscle ( p <0.05) than in the supraspinatus muscle ( average value of middle measuring site 51.4 compared to 35.0). Furthermore, the number of spatially connected and homogeneous regions ( blobs) was higher in the vastus lateralis ( p <0.05) than in the supraspinatus ( average for m. vastus lateralis: 0.092 mm(-2) and for m. supraspinatus: 0.016 mm(-2)). Conclusion: The higher intensity and the higher number of blobs in the vastus lateralis muscle indicates that the thigh muscle contained more non-contractile components than the supraspinatus muscle, and that the muscle was coarser. The image analyses supplemented each other and gave a more complete description of the tissue composition in the muscle than the mean grey-scale value alone.