Spinal anaesthesia is generally preferred for Caesarean section. Its superiority for the baby is often assumed. Umbilical artery acid-base status provides a valid index of fetal welfare. Twenty-seven studies reporting neonatal acid-base data with different types of anaesthesia were used to compare umbilical artery or vein pH and base deficit, using random-effect meta-analysis. Cord pH was significantly lower with spinal than with both general (difference: -0.015; 95% CI -0.029 to -0.001; 13 studies, 1272 subjects) and epidural anaesthesia (difference -0.013; 95% CI -0.024 to -0.002; 11 studies, 828 subjects). Larger doses of ephedrine contributed to the latter effect (p = 0.023). Sixteen studies reported a base deficit, which was significantly higher for spinal than for general (difference 1.109; 95% CI 0.434-1.784 mEq.l(-1); seven studies, 695 subject) and epidural anaesthesia (difference 0.910; 95% CI 0.222-1.598 mEq.l(-1); seven studies, 497 subjects). Spinal anaesthesia cannot be considered safer than epidural or general anaesthesia for the fetus.