Our objective was to examine the neonatal outcome of second twins depending on presentation and mode of delivery. Using a database we analyzed the short-term neonatal outcome in twin pregnancies offered a trial of labor with special emphasis on the second twin depending on presentation and mode of delivery. Neonatal outcome was evaluated by Apgar scores, umbilical cord blood pH values, and perinatal or neonatal morbidity and mortality. Overall, in 219 (78%) of 281 pregnancies successful vaginal birth (VB) of both twins (VB–VB) was possible, 48 (17%) women had to be delivered by cesarean section (CS) of both twins (CS–CS), and in 14 (5%) women the second twin had to be delivered by CS after VB of the first twin (VB–CS). Successful VB was most common for vertex-vertex (V/V; n = 171, 82%) and vertex-nonvertex (n = 48, 75%) presentation (V/NV). Twins delivered by VB–CS had the lowest values for pHart (p = .006) and pHven (p = .010). pHart less than or equal to 7.00 values occurred only in second twins delivered VB–VB or VB–CS. Lower Apgar scores of the second twin occurred more frequently in the VB–CS and in the VB–VB than in the CS–CS groups (ps < .05). Lower levels of pHart (p = .002) and frequency of pHart less than or equal to 7.00 occurred more often in nonvertex second twins than in vertex second twins (p < .022). The high CS rate in V/NV presentation and the significantly worse perinatal short-term outcome of NV second twins after VB of the first twin underline that randomized studies are necessary to evaluate the best delivery mode for V/NV twins.