OBJECTIVE: Placenta Accreta Spectrum (PAS) disorders have become a major iatrogenic obstetric complication worldwide. Data on the accuracy of ultrasound examination diagnosis are limited by incomplete confirmation and variability in the description of the different grades of PAS at delivery. The aim of this study was to compare our prenatal routine sonographic screening and diagnostic scoring system with a standardized clinical grading system at birth in patient at risk of PAS. STUDY DESIGN: This is a retrospective cohort study of 607 pregnant patients with at least one prior cesarean delivery between December 2013 and December 2018. All patients were assessed for PAS using our institutional prenatal sonographic scoring system and the corresponding ultrasound findings were compared with those of a standardized clinical intra-operative macroscopic grading system of the degree of accreta placentation at vaginal birth or laparotomy. RESULTS: PAS was diagnosed clinically at birth in 50 (8.2%) cases, 17 of which were confirmed by histopathology. A low (score ≤ 5), medium (score 6-7), high (score ≥ 8) probability for PAS was reported in 502, 61 and 44 cases, respectively. The probability score increased significantly (p < .001) in women ≥2 prior cesarean deliveries, with an anterior low-lying/placenta previa, with absent clear space, increased in retroplacental vascularity and with the size and numbers of lacunae. The number of cases classified clinically as grade 1 (non-PAS) and 3 (adherent PAS) was significantly (p < .001) lower in women with a high probability score whereas the rates of the other grades was significantly (p < .001) higher. The widest discrepancy between ultrasound probability score and clinical grade was found for grade 2 which, describes a partial placental adherence and grades 4 and 5 which, refer to placental percreta which describes tissue having invade trough the uterine serosa and beyond. CONCLUSIONS: Both ends of the spectrum of accreta placentation remain difficult to diagnose antenatal and clinically at birth, in particular when no histopathologic confirmation is available. There is a need to develop ultrasound accuracy score systems that can differentiate between the different grades of PAS and which are validated by standardized clinical and pathology protocols.