The aim was to determine the diagnostic performance of 3-dimensional virtual reality ultrasound (3D_VR_US) and conventional 2- and 3-dimensional ultrasound (2D/3D_US) for first-trimester detection of structural abnormalities. Forty-eight first trimester cases (gold standard available, 22 normal, 26 abnormal) were evaluated offline using both techniques by 5 experienced, blinded sonographers. In each case, we analyzed whether each organ category was correctly indicated as normal or abnormal and whether the specific diagnosis was correctly made. Sensitivity in terms of normal or abnormal was comparable for both techniques (P = .24). The general sensitivity for specific diagnoses was 62.6% using 3D_VR_US and 52.2% using 2D/3D_US (P = .075). The 3D_VR_US more often correctly diagnosed skeleton/limb malformations (36.7% vs 10%; P = .013). Mean evaluation time in 3D_VR_US was 4:24 minutes and in 2D/3D_US 2:53 minutes (P < .001). General diagnostic performance of 3D_VR_US and 2D/3D_US apparently is comparable. Malformations of skeleton and limbs are more often detected using 3D_VR_US. Evaluation time is longer in 3D_VR_US.