The diagnosis of CDH in infant is often difficult. The clinical tests for looseness of the hip are not always reliable and findings observed on plain radiographs of the hip are apt to be misinterpretated in infants, especially younger than 3 months of age. At the present time, real-time Ultrasonography (US) is the only readily available noninvasive imaging modality that can directly image the cartilaginous femoral head and its relationship to the bony and cartilaginous acetabulum. 129 hips of 37 patients with CDH and 12 normal control neonates were evaluated with US. In order to determine the hip instability, we devised Hip Coverage Index (HCI), which represents the percentage of femoral head that is covered under the acetabulum, measured on both coronal image and transverse image in neutral position. In 14 hips the effectiveness of treatment by harness or brace were evaluated longitudinally with serial US follow-up study. The hips were divided into three groups: Normal (82), Dysplasis Subluxation (19), and Dislocation (14). There were 12 boys and 37 girls. The mean age was 6 months, ranging from 2 days to 21 months. The mean Hip Coverage Index (HCI) was 61.0±7.6% in normal control and contralateral hips, as compared with 43.1±5.9% in Dysplasia Subluxation and 25.1±10.7% in Dislocation (P<0.05). HCI was correlated well with the indices (α, β angle) of Graf's method. The index increased with time and paralleled the degree of improvement of stability in the group treated by harness or brace. HCI was proved to be an easily reproducible method to determine the severity of hip instability and the effectiveness of treatment.