The mainstay of dyspnea palliation remains altering its central perception. Morphine is the main drug and anxiolytics have a less established role. This trial assessed the role of midazolam as adjunct therapy to morphine in the alleviation of severe dyspnea perception in terminally ill cancer patients. One hundred and one patients with severe dyspnea were randomized to receive around-the-clock morphine (2.5 mg every 4 hours for opioid-naïve patients or a 25% increment over the daily dose for those receiving baseline opioids) with midazolam rescue doses (5 mg) in case of breakthrough dyspnea (BD) (Group Mo); around-the-clock midazolam (5 mg every 4 hours) with morphine rescues (2.5 mg) in case of BD (Group Mi); or around-the-clock morphine (2.5 mg every 4 hours for opioid-naïve patients or a 25% increment over the daily dose for those receiving baseline opioids) plus midazolam (5 mg every 4 hours) with morphine rescue doses (2.5 mg) in case of BD (Group MM). All drugs were given subcutaneously in a single-blinded way. Thirty-five patients were entered in Group Mo, 33 entered in Mi, and 33 entered in MM. At 24 hours, patients who experienced dyspnea relief were 69%, 46%, and 92% in the Mo, Mi, and MM groups, respectively (P = 0.0004 and P = 0.03 for MM vs. Mi and MM vs. Mo, respectively). At 48 hours, those with no dyspnea relief (no controlled dyspnea) were 12.5%, 26%, and 4% for the Mo, Mi, and MM groups, respectively (P = 0.04 for MM vs. Mi). During the first day, patients with BD for the groups Mo, Mi, and MM were 34.3%, 36.4%, and 21.2%, respectively (P = NS or not significant), whereas during the second day, these percentages were 38%, 38.5%, and 24%, respectively (P = NS). The data demonstrate that the beneficial effects of morphine in controlling baseline levels of dyspnea could be improved with the addition of midazolam to the treatment.