The efficacy of neurolytic coeliac plexus block (NCPB) guided by computerized tomography (CT) was compared with pharmacological therapy in the treatment of pain due to pancreatic cancer. The study involved 56 patients who were placed randomly in either a NCPB group and pharmacological therapy group. At day 1, 7, and 14, the visual analogue scale (VAS) pain scores of the NCPB group were significantly lower than those of the pharmacological therapy group (P < 0.01), with values of 1.3 +/- 0.8 versus 4.1 +/- 0.9, 1.7 +/- 1.1 versus 3.1 +/- 1.1, and 2.0 +/- 1.1 versus 2.9 +/- 0.6, respectively. However, the differences in the improvement of quality of life (QOL) between two groups were not statistically significant. Moreover, the dose of opioid was significantly lower in the patients of group 1 than those of group 2, while the complications related to NCPB were transient. We therefore concludes that CT-guided NCPB with alcohol is an effective and safe modality in the management of intractable pancreatic cancer pain.