The aim of this study is to evaluate the clinical risk factors, accuracy of antenatal ultrasound for diagnosis, and the effect of these on pregnancy outcome. It is a retrospective study looking at cases which had hysterectomy following vaginal or caesarean section deliveries from 1993 to 2005. Data regarding the maternal demographic characteristics, number of previous CS, number of previous termination/curettage, antenatal scan findings (state features) and the gestation at which accreta was first suspected/diagnosed, MRI scan findings, pregnancy outcome (need for hysterectomy, amount of blood loss, amount of transfusion, length of ICU and hospital stay, other maternal complications, and neonatal outcome) were collected and evaluated. There were a total of 40 cases diagnosed to have abnormal placental attachment and majority of these were actually diagnosed antenatally by sonography. Visualisation of an absence or thinning of hypoechoic myometrial zone had the highest sensitivity to detect placenta accreta followed by intraplacental lacunae, focal mass tissue elevation and disruption of uterine serosal bladder wall.