To explore the impact of epidural analgesia on maternal and neonatal outcomes, especially the relation between epidural analgesia and intrapartum fever. A retrospective cohort study was conducted in a tertiary hospital for all deliveries from November 2017 to December 2017. A total of 506 women were divided into epidural and non-epidural group by whether to receive analgesia or not. Univariate and multivariate analyses were performed with P < 0.05 as significant. Epidural analgesia was associated with higher risk of maternal intrapartum fever (relative risk [RR] = 3.28, 95% confidence interval, 1.55-6.95), more intravenous use of antibiotics (36.66% vs 17.04%, P<0.001), longer time of second stage (58.55 ± 33.75 vs 47.39 ± 28.36 min,P = 0.001) and longer total duration of labor (790.32 ± 433.71 vs 461.33 ± 270.39 min,P<0.001), but had no influence on mode of delivery, the amount of post-partum hemorrhage or hospital stay after delivery and all the neonatal outcomes we studied. Further time effect analysis found that epidural analgesia less than 6 h did not increase the risk of intrapartum fever (RR = 1.73, P = 0.15), however, when epidural analgesia lasted over 6 h, it significantly increased the risk of fever (RR = 5.23, P<0.001) but did not increase more adverse outcomes. Having epidural anesthesia 6 h or more increases the risk of developing fever, but the prognosis of mothers and children is less affected.