Background. Ultrasound (US) is routinely used for hip screening in children with developmental hip disorder, whereas surveillance in children with cerebral palsy (CP) is based on repeated X-ray assessments. ----- Objective. To evaluate US as a diagnostic tool in screening for decentered hips in children with CP. ----- Patients and Methods. Prospective, diagnostic assessor-blind study. Consecutive CP patients (age 2-8 years) with severe motor disability underwent US (lateral longitudinal scan to determine lateral head distance, LHD) and X-ray hip assessment (migration percentage, MP). Diagnostic properties of LHD in detecting hips with MP≥33% were evaluated overall (n=100) and for hips assessed at the age 24-60 (n=38) or 60< to ≤96 months (n=68). Fifty hips underwent US assessment by two investigators to evaluate inter-rater reliability and agreement. ----- Results. Prevalence of MP≥33% was 22.0% overall and 26.2% and 19.4% in the two age-based subsets, respectively. LHD well discriminated hips with MP≥33% (areas under the ROC curves 94%, 99% and 92%, respectively). At the optimum cut-off values of LHD (5.0, 5.0 and 4.8 mm, respectively), sensitivity was 95.5% overall and 100% in the two age-based subsets, whereas specificity was 85.9%, 96.4%, 72.0%, respectively. Consequently, positive predictive value was relatively low, but negative predictive value was 98.5% (95% CI 92.1-100) overall and 100% in the two age-based subsets. Inter-rater reliability was high (intraclass correlation coefficient= 0.98, 95%CI 0.97-0.99) and 95% limits of agreement were reasonably narrow (-1.203 mm to 0.995 mm). ----- Conclusion. In children with CP, US can be reliably used in screening for decentered hips and can greatly reduce the need for repeated X-rays.