PURPOSE: To prospectively compare three-dimensional (3D) ultrasonography (US) and magnetic resonance (MR) imaging in the assessment of lung volumes in fetuses with congenital diaphragmatic hernia. MATERIALS AND METHODS: Informed consent was obtained for this ethics committee-approved study. Left and right lung volumes were measured by using the 30 degrees virtual organ computer-aided analysis 3D US technique and a transverse multiplanar T2-weighted MR imaging technique in 43 fetuses with isolated congenital diaphragmatic hernia. Regression analysis was used to determine the significance of the association between the two methods. RESULTS: The 43 fetuses were assessed in a total of 78 examinations. Median gestational age at the examinations was 28 weeks (range, 18-38 weeks). In all examinations, it was possible to visualize and measure both the ipsilateral and the contralateral lungs with MR imaging. In contrast, with 3D US, the contralateral lung could be measured in all examinations, but the ipsilateral lung could be measured in only 44 (56%) examinations. For the contralateral lungs, there was a significant association between 3D US and MR imaging measurements (r = 0.86, P < .001). Although the mean lung volume measured with 3D US was 25% lower than that measured with MR imaging, the ratio of observed volume to expected normal mean volume for gestation was not significantly different between the two methods (3D US, 0.48; MR imaging, 0.52). In the 44 examinations in which the ipsilateral lung could be measured with both methods, 3D US volumes were not significantly different from MR imaging volumes, and the association was weaker (r = 0.39, P < .05) in the ipsilateral lungs than in the contralateral lungs. CONCLUSION: For congenital diaphragmatic hernia, 3D US provides a reliable measurement of the contralateral but not the ipsilateral lung.