Our objective was to determine the perinatal outcome of first- and second-born twins compared to singletons, born at the same gestational age. To that end we conducted a case-control study in Flanders (Northern Belgium). During a 10-year period (01.01.1999–31.12.2008), the entire twin population — 11,154 first- and 11,118 second-born twins (cases) — was compared to 22,228 singletons (controls) with respect to fetal and neonatal (0–27 days) mortality. Only case and control infants of ≥ 500 grams were included, which explained the unequal number of first- and second-born twins. Mothers and their infants of cases and of controls were derived from the Flemish perinatal database and were matched for maternal age and parity, gestational age and gender of the offspring. The main outcome measures were fetal and neonatal mortality according to gestational age. The frequency of fetal death was statistically significantly less frequent in preterm born twins than in singletons, except at term where the reverse was seen in second-born twins compared to controls. After adjustment for congenital malformations, the results stayed unchanged. Below 28 weeks gestation, singletons had a significantly lower neonatal mortality rate than twins that persisted after adjustment for congenital malformations: the first-born twin versus singleton OR 1.71 (1.17–2.51) and second-born versus singleton OR 2.09 (1.43–3.05). Between 28 and 32 weeks, the second-born twin showed a survival advantage over the control singleton. Between 32 and 36 6/7 weeks both twins had a significantly higher survival rate than the corresponding singleton controls. However, after adjustment for congenital malformations, the aforementioned differences between 28 and 36 6/7 weeks disappeared. When at term, twins and singletons had a comparable, though very low, neonatal death rate. These results confirm previous published data. In conclusion, we demonstrated that the neonatal death rate was lower for twins between 32 and 36 weeks (from 28 weeks for the second born twin) when compared to a singleton of the same gestational age. After adjusting for congenital malformations, there was no statistical significant difference.