Clinicians, scientists & radiologists developing obstetric MRI. Uterine blood flow (UBF) is an important biomarker of fetal well-being but not directly measured during routine ante-natal ultrasound examinations because accurate bulk flow measurements with Doppler ultrasound are difficult to obtain and prone to multiple errors. Indirect markers of materno-placental vascular resistance such as the pulsatility and resistive indices are considered more reliable for identifying abnormalities and guiding clinical management. MRI is becoming more widely used in obstetrics and offers several methods of non-invasively quantifying blood flow. ASL & IVIM techniques¹ have been reported but do not provide absolute flow measurements and are not widely available. Early attempts with phase contrast proved unsuccessful owing to difficulty identifying the uterine arteries². Recently we demonstrated, using a combination of multiplanar balanced SSFP imaging and cine phase contrast information, a technique³ that identified the uterine arteries in the majority of normal subjects in the early third trimester. The aim of this work is to use the same approach to estimate total uterine blood flow with cine phase contrast and determine if this correlates with estimated fetal weight and actual birthweight. This study was ethically approved and 35 participants with a normal singleton pregnancy at 28-32 weeks gestation were recruited. They underwent MRI examination in a left decubitus position using an 8 channel cardiac array and a 1.5T MRI system (DV450 GEHC, Waukesha, WI). Following sagittal and axial FIESTA breath-hold imaging through the uterus, an oblique coronal image plane was positioned immediately superior to, and parallel with, the external iliac arteries. A cardiac gated cine phase contrast study was performed in this plane using the following parameters: TR/TE 6.45/3.1msec, slice thickness 7mm, FOV 36cm, matrix 192x256, flip angle 30, 60 phases, Venc 80-90cm/sec with S/I encoding and acq time of 90-120 seconds depending on heart rate. Using colocation with the Fiesta and phase contrast images, vessels in the uterine wall that passed orthogonally through the image plane, and with flow towards the uterus, were identified as likely uterine arteries (Figure 1) and manual ROIs were placed around each artery and adjacent background tissue. An in-house flow analysis program written in Matlab was used to unwrap aliasing, provide background correction, and calculate mean bulk flow in mls/minute. Total UBF was the sum of estimated flows in left and right uterine arteries. Participants were excluded if at least one artery on each side was not identified, if the analysis program failed to unwrap aliasing or if there were severe artefacts eg from motion. Estimated fetal weights were obtained from same day routine ultrasound examinations and actual birthweights were obtained from subsequent obstetric records.At MRI 4 multiple uterine arteries on one side were identified, 1 artery had severe aliasing and 3 vessels could not be confidently identified, leaving 66 vessels in 31 patients for analysis. In 1 patients the birthweight could not be obtained owing to the participant moving country. The total UBF ranged from 225 to 1014 mls/minute with a mean of 521.8 mls/minute. There was quite wide variation between left and right uterine flow in the same patients (absolute differences ranged from 5 to 479, mean 124.4 mls/min). There was no correlation with estimated fetal weight (r = -0.12, Figure 2) however UBF did correlate moderately with actual birthweight (r = 0.3, Figure 3).The UBF results obtained using cinePC MRI are in the same range as those obtained with careful Doppler ultrasound examinations⁴ and comparable with Fick principle studies performed at term in the 1950s using radioactive sodium⁵ and nitrous oxide. We speculate that the variation of gestational age in the sample may account for the lack of correlation of EFW with UBF. Other studies have also demonstrated a relationship between measures of placental blood flow and pregnancy outcomes such as birthweight¹. This study used relatively low spatial resolution cinePC which could be further optimised to reduce partial volume artefacts and potentially increase UBF measurement accuracy.This study demonstrates in the early third trimester that the uterine arteries can be identified at MRI, and that cinePC MRI of these vessels is feasible and gives uterine blood flow results comparable with other methods. Cine PC is widely available on commercial MRI systems and with careful technique could prove a more reliable and practical technique for non-invasive UBF measurements than Doppler ultrasound.