Measurement is one of the key parameters of US examination and represents an essential part for a correct interpretation of ultrasound (US) images and for distinguishing normal from pathologic conditions. In the musculoskeletal system, US measurements involve calculation of linear distance, area, or volume. Distance measures are the most commonly used by far. They are usually obtained on the freezed image frame by moving a visible cursor on the screen via a track ball: the process may take a couple of seconds. The precision of distance measurements reached by current technology and high frequency broadband transducers is very high, and, in the best circumstances, spot reflectors measuring 0.1 mm in size can be resolved as separate structures. Area measurements are used in more specific settings. Because most of area measurements relate to round/oval structures, built-in systems that produce an ellipse on the screen can give results quickly by adjusting size and shape on the structure to be evaluated with the track ball. On the other hand, freehand area measurements need a steady hand: it may be difficult to draw on the screen a trace that perfectly overlaps the structure of interest. However, positive and negative errors produced by the caliper spots that deviate from the intended position on either side of the outline, at the end, tend to balance. Automated recognition algorithms have recently been introduced to make precise tracing around a given structure without wasting time. If implemented in the equipment software, volume algorithms usually refer to ellipsoid structures or recall obstetrical shapes. In general, these systems do not perfectly fit for use in the musculoskeletal system, where skeletal muscles have variable conformation. The advent of 3D and 4D ultrasound is opening new perspectives in this field. One of the main drawbacks of linear array transducers is the limited extension of the field of view that makes measurements of elongated structures in the musculoskeletal system impractical. Thus, spatial relationships and sizes in the US images often must be synthesized in the mind of the sonologist from multiple real-time images that display only portions of the relevant anatomy (Lin et al. 1999). With extended field-of-view systems, however, geometric measurements can be obtained effectively from lesions larger than the field of view of the transducer with <5% error (Fig. 1) (Fornage et al. 2000).