Background: Methadone is used more and more as a second-line treatment for refractory cancer pain in palliative care patients. Methods: The study aimed to compare the effect of two methadone titration methods (Stop and Go vs progressive titration) in 146 palliative care cancer patients with pain inadequately relieved or intolerant to level 3 opioids. The primary endpoint was the rate of success/failure at Day 4 defined by pain relief (reduction of at least two points of the numerical scale (0 - 10) AND a pain score < 5 for 2 consecutive days) AND no overdose (Rudkin scale ≥ 3 AND respiratory rate < 8/min). Results: Pain was nociceptive in 16% and mixed in 84%. 85% of the patients had breakthrough pain. Half received oxycodone, 1/5 fentanyl, 1/5 morphine and < 10% hydromorphone. Reasons for switching were lack of efficacy isolated (56%), or with intolerance to previous opioid (38%). More than 2/3 of the patients reached a pain score < 5 for two days at D4, adequate pain relief was obtained in 80% of the patients (median of 3 days in both groups (p = 0.12)) and lasted until D56. The rate of success/failure was about 40% at D4 with no difference between the two methods. Overdoses defined as respiratory rate < 8/min were observed in 13.2% patients throughout the study with no difference between groups. 9 required naloxone. The two methods were equally considered easy by about 60% of the clinicians. Conclusions: Methadone is an effective and sustainable second-line alternative opioid in the treatment of cancer pain in palliative care patients inadequately relieved or intolerant to level 3 opioids. The two methods of titration of methadone are comparable in terms of efficacy, safety and ease of use.