INTRODUCTION: Pain is a common symptom in cancer patients even when analgesics were given. OBJECTIVE: To assess the prognostic value of intravenous infusion of non-selective purine receptors agonist adenosine triphosphate for effectiveness of anticonvulsant gabapentin in cancer patients with moderate and severe pain who received non-selective inhibitor of cyclooxygenase plus weak opioid tramadol. MATERIALS AND METHODS: Thirty-four cancer patients with intensive pain were scheduled for the study. The intravenous infusion of adenosine triphosphate 35–45 mg∙kg–1∙min–1 was performed within the period from 100 to 160 minutes. Then all patients were taken gabapentin (900 mg in three times daily). Pain was estimated 30 minutes before infusion, 30 minutes after infusion, and after taking 900 mg/day gabapentin for 4 days. Moreover, we studied prognostic significance of adenosine triphosphate infusion for the effectiveness of gabapentin administration. RESULTS: We revealed significant reduction of pain intensity after adenosine triphosphate infusion (Z = 4.0; р < 0.0001 — Wilcoxon signed rank test). The same result was obtained after taking of gabapentin for 4 days (Z = 4.9; р < 0.0001 — Wilcoxon signed rank test). Moreover, we found statistically moderate correlation link (t(N – 2) = 3.94; closeness correlation = 0.57; р < 0.0004 — Spearman’s rank correlation coefficient) between pain intensity value after adenosine triphosphate infusion and taking of gabapentin. Regression analysis demonstrated satisfactory predictive ability of the resulting model (R2 = 0.55 (corrected R2 = 0.53); F = 38.74; р < 0.0001). CONCLUSIONS: 1. Intravenous infusion of adenosine triphosphate may has significance for prognosis of taking anticonvulsant gabapentin effectiveness in cancer patients with moderate and severe pain who received non-selective cyclooxygenase inhibitor plus weak opioid tramadol. 2. Intravenous infusion of adenosine triphosphate or taking anticonvulsant gabapentin may significantly reduce pain intensity in cancer patients who had weak effect of administration of non-selective inhibitor of cyclooxygenase plus weak opioid tramadol.