More than 85% of cancer-related pain is pharmacologically controllable, but some patients require interventional treatments. Although audit assessment of these interventions is of importance to clarify the types of patients likely to receive benefits, there have been no multicenter studies in Japan. The primary aims of this study were (1) to clarify the frequency of neural blockade in certified palliative care units and palliative care teams, (2) determine the efficacy of interventions, and (3) explore the predictors of successful or unsuccessful intervention. All patients who received neural blockade were consecutively recruited from seven certified palliative care units and five hospital palliative care teams in Japan. Primary responsible physicians reported pain intensity on the Support Team Assessment Schedule, performance status, communication levels on the Communication Capacity Scale, presence or absence of delirium, opioid consumption, and adverse effects before and one week after the procedure on the basis of retrospective chart review. A total of 162 interventions in 136 patients were obtained, comprising 3.8% of all patients receiving specialized palliative care services during the study period. Common procedures were epidural nerve block with local anesthetic and/or opioids (n = 84), neurolytic sympathetic plexus block (n = 24), and intrathecal nerve block with phenol (n = 21). There were significant differences in the frequency of neural blockade between palliative care units and palliative care teams (3.1% vs. 4.6%, respectively, P = 0.018), and between institutions whose leading physicians are anesthesiologists or have other specialties (4.8% vs. 1.5%, respectively, P < 0.001). Pain intensity measured on the Support Team Assessment Schedule (2.9 +/- 0.8 to 1.7 +/- 0.9, P < 0.001), performance status (2.7 +/- 1.0 to 2.4 +/- 1.0, P < 0.001), and opioid consumption (248 +/- 348 to 186 +/- 288 mg morphine equivalent/day, P < 0.001) were significantly improved after interventions. There was a tendency toward improvement in the communication level measured on the Communication Capacity Scale. There was no significant improvement in the prevalence of delirium, but six patients (32%) recovered from delirium after interventions. Adverse effects occurred in 9.2%, but all were predictable or transient. No fatal complications were reported. Pain intensity was significantly more improved in patients who survived 28 days or longer than others (P = 0.002). There were no significant correlations of changes in pain intensity with the performance status or previous opioid consumption. In conclusion, neural blockade was performed in 3.8% of cancer patients who received specialized palliative care services in Japan. Neural blockade could contribute to the improvement of pain intensity, performance service status, and opioid consumption without unpredictable serious side effects.