Current ultrasound techniques can accurately determine the chorionicity of twins, but not zygosity. We previously proposed that the zygosity of spontaneously conceived twins can be determined at early ultrasound, where 2 corpora lutea infers dizygosity, and 1 implies monozygosity. Here we did a case series, comparing zygosity predicted using this method with definitive DNA genotyping of twins after birth. We retrospectively identified 14 ultrasound reports of spontaneous twin pregnancies at 6(+0 days) to 13+6 weeks' gestation, where both ovaries were seen and the number of corpora lutea documented. We visited all twin pairs, obtained buccal smears, and determined zygosity by genotyping 9 independent microsatellite markers. All 8 cases where 2 corpora lutea were seen were dizygotic pregnancies. One further case where 3 corpora lutea were seen was also dizygotic. All 3 sets of monozygotic twins had 1 corpus luteum. There were 2 cases incorrectly assigned, where 1 corpus luteum was seen in dizygotic pregnancies. We conclude if 2 corpora lutea are seen at a first trimester ultrasound of spontaneously conceived dichorionic twins, they appear to be almost certainly dizygotic. However, if 1 corpus luteum is seen in dichorionic twins, zygosity cannot be determined with certainty since it is either monozygotic, or dizygotic where a second corpus luteum has been missed.