Nitrous oxide at 50% inhaled concentration has been shown to improve depressive symptoms in patients with treatment-resistant major depression (TRMD). Whether a lower concentration of 25% nitrous oxide provides similar efficacy and persistence of antidepressant effects while reducing the risk of adverse side effects is unknown. In this phase 2 clinical trial (NCT03283670), 24 patients with severe TRMD were randomly assigned in a crossover fashion to three treatments consisting of a single 1-hour inhalation with (i) 50% nitrous oxide, (ii) 25% nitrous oxide, or (iii) placebo (air/oxygen). The primary outcome was the change on the Hamilton Depression Rating Scale (HDRS-21). Whereas nitrous oxide significantly improved depressive symptoms versus placebo (P = 0.01), there was no difference between 25 and 50% nitrous oxide (P = 0.58). The estimated differences between 25% and placebo were −0.75 points on the HDRS-21 at 2 hours (P = 0.73), −1.41 points at 24 hours (P = 0.52), −4.35 points at week 1 (P = 0.05), and −5.19 points at week 2 (P = 0.02), and the estimated differences between 50% and placebo were −0.87 points at 2 hours (P = 0.69), −1.93 points at 24 hours (P = 0.37), −2.44 points at week 1 (P = 0.25), and −7.00 points at week 2 (P = 0.001). Adverse events declined substantially with dose (P < 0.001). These results suggest that 25% nitrous oxide has comparable efficacy to 50% nitrous oxide in improving TRMD but with a markedly lower rate of adverse effects.