a b s t r a c tObjectives: To test the hypothesis that third trimester placental biometry and volume can be measured by two-dimensional (2D) and three-dimensional (3D) ultrasound in utero, determining which method of measurement was most strongly correlated with true placental size ex vivo. Methods: Singleton pregnancies underwent placental ultrasound within seven days of delivery (n ¼ 87, 29 þ3 e41 þ5 weeks). Length and width (linear and curvilinear) and depth were estimated. Placental volume (PV) was estimated using 2D ellipse and shell techniques and 3D rotational (15⁰ and 30⁰ rotation angles) and multiplanar (5 and 10 mm slicing intervals) techniques. Measurements were compared to their true correlates following delivery. Intra-and inter-observer reliabilities of candidate placental size estimates were assessed by intraclass correlation coefficient (ICC). Results: Curvilinear placental length (R s ¼ 0.24, p ¼ 0.031), width (R s ¼ 0.27, p ¼ 0.013) and depth (R s ¼ 0.31, p ¼ 0.0056) correlated well with ex vivo measurements. All methods of PV estimation were related to ex vivo volume (R s ! 0.32, p < 0.01) but not placental weight (p > 0.05); 30 rotational estimation demonstrated the strongest biological correlation (R s ¼ 0.40, p ¼ 0.0004). Intra-and interobserver placental size measurements intraclass correlation coefficients were suboptimal (0.59e0.70 and 0.10e0.58 respectively). Discussion: We have demonstrated that it is possible to obtain information about the size of the third trimester placenta in utero using 2D and 3D ultrasound. However it is essential that the reliability (particularly interobserver reliability) of these estimates is improved prior to prospective studies to determine their predictive value.