Lumbar epidural analgesia (LEA) is commonly used for labor analgesia but up to 13% of epidural catheters fail and require replacement. Combined spinal-epidural analgesia is associated with a lower catheter failure rate. Few data exist regarding catheter replacement rates after dural-puncture epidural (DPE). We conducted a retrospective analysis comparing catheter failure rates between epidural and DPE techniques. This retrospective single-center trial reviewed all labor neuraxial analgesia procedures among 18 726 women across five years, and identified 810 DPE and 2667 LEA procedures. Catheter failure rates, consisting of replacement or requirement of general anesthesia for cesarean delivery, were compared. Propensity score matching was used to balance the groups. Dural-puncture epidural was associated with significantly fewer catheter failures compared with LEA (74/759 vs. 49/759, odds ratio 0.64, 95% CI 0.44 to 0.93, P=0.02). Sensitivity analysis excluding cases of general anesthesia confirmed this relationship. Risk factors identified for catheter failure included age, body mass index, and nulliparity. Dural-puncture epidural was associated with a longer mean time to catheter replacement (918 min vs. 609 min, P=0.04). Kaplan-Meier and Cox multivariate analyses confirmed this relationship. There was no significant difference in the requirement for epidural analgesia supplementation, but DPE required supplementation significantly later than LEA. There was no difference in the rate of headache or epidural blood patch between groups. Dural-puncture epidural is associated with fewer catheter failures and replacements than LEA, without an increase in the rate of post-dural puncture headache or epidural blood patch.