Supplementary doses of opioids are recommended to relieve dyspnea in terminally ill cancer patients. We conducted a randomized continuous sequential clinical trial to evaluate their efficacy. We recruited 33 terminally ill cancer patients from three palliative care centers, all of whom had persistent dyspnea after rest and treatment with oxygen. Patients formed 15 successive pairs matched on route of administration. Within each pair, the order of allocation was randomly assigned, one patient receiving 25%, the other 50% of his 4-hourly opioid dose. Five measurements of dyspnea intensity and respiratory frequency were made during 4 hours of follow-up. For each pair, a preference was attributed to the more effective regimen. The two regimens received an almost equal number of paired preferences (8 vs. 7). Overall, both mean dyspnea intensity and respiratory frequency decreased significantly relative to baseline. Dyspnea reduction was relatively greater in patients with initially low and moderate dyspnea intensity. In terminally ill cancer patients with persistent dyspnea, 25% of the equivalent 4-hourly dose of opioid may be sufficient to reduce both dyspnea intensity and tachypnea for 4 hours.