Background. Prevention of loss of containment has become an accepted principle in the treatment of Perthes' disease. The pre-requisite is early recognition. It is based on evaluation of plain radiographs and more recently, on the study of Magnetic Resonance (MR) images which allow discrimination of early cartilaginous changes. Ultrasonography (US) allows visualisation of the lateral cartilaginous portion of the femoral head and the acetabular rim including the labrum and measurement of femoral head protrusion/lateralisation. The purpose of this paper is to highlight its potential for monitoring of containment . Materials and methods. We present typical MR and US images to demonstrate the anatomic landmarks of the normal hip joint and to define the parameters of protrusion in Perthes' disease. We selected three illustrative cases that had undergone routine imaging of both hip joints by MR imaging and ultrasound for evaluation of containment. Radiographs of the hips were also available. In radiographs we assessed the coverage of the femoral head, i.e. containment, by the well established Acetabulum-Head Index (AHI) and in MR imaging by the Cartilaginous Acetabulum-Head Index (CAHI). In US we assessed the uncoverage, i. e. protrusion, by the Lateral Cartilage Distance (LCD). Changes in the important morphological MR containment features were also noted. Results. There was a significant increase in the LCD in all Perthes hips (6.2, 7.4, 11.6 mm) when compared to the unaffected side (5.2, 5.1, 4.1 mm) and also when compared to the published mean normal value (5.4 +/- 0.9 mm). Correspondingly, the CAHI values were significantly decreased (75, 69, 67% versus 87, 79, 81%), also in comparison to the published limits (77, 75, and 73% respectively). As for the AHI only the value of 71 % in the third case represented a definite decrease below published normal limits (86 and 80.7% respectively). In the 1st case we diagnosed adequate containment, in the 2nd containment at risk, and in the 3rd loss of containment. In the 2nd case the AHI of 90 % suggested adequate containment whereas considerable protrusion/lateralisation was evident in MR imaging and US. The CAHI was only 69%. It showed that assessment by plain radiographs is less reliable because the cartilaginous portion of the hip joint is not included in interpretation. We were able to demonstrate a good agreement between LCD and CAHI in our cases. Conclusion. US can be helpful for monitoring of containment in Perthes' disease allowing a closer follow-up and a reduction of serial radiographs and MR exams.