Background: The development of back pain following epidural analgesia is one reason for patient refusal of neuraxial analgesia. The primary endpoint of this study was to compare the incidence and severity of back pain following midline and paramedian epidural technique. The secondary endpoint was to identify the risk factors associated with the occurrence of back pain.Methods: This prospective randomized study included 114 patients receiving thoracic epidural catheterization for pain management following upper abdominal or thoracic surgery. Patients were allocated to either the midline or the paramedian group by computer-generated randomization. An investigator who was blinded to the patient group interviewed patients at 24, and 48 h, and 3–5 days after surgery about the existence of back pain and its severity.Results: The total incidence of back pain following epidural anesthesia was 23.8% in the midline group and 7.8% in the paramedian group. The numerical rating scale of back pain was not different between the two groups at 24 h and 4 days after surgery. The paramdian technique was associated with a lower incidence of back pain than the midline technique (95% confidence interval 0.05–0.74, odds ratio 0.2, P < 0.01). However, the number of attempts, surgical position, body mass index, and duration of surgery were not associated with back pain.Conclusions: This study showed that the midline group of thoracic epidural analgesia demonstrated higher incidence of back pain than the paramedian group. However, the pain was mild in intensity and decreased with time in both groups.