INTRODUCTION: Several methodologies have been introduced to reduce oxygen (O2) usage during home O2 therapy. Pulse-dose O2 therapy has been reported to reduce O2 usage and costs. PURPOSE: Our purpose was to compare the clinical effectiveness and cost of pulse-dose O2 therapy to continuous-flow O2 therapy via nasal cannula in patients with chronic obstructive pulmonary disease (COPD), who were living at home. METHODS: We analyzed arterial blood samples drawn from 20 patients with hypoxemia on room air (mean PaO2 52 torr) while they were using continuous-low O2 at 2 L/min and while they were using pulse-dose O2 at a dose of 33 mL/pulse. In addition, total O2 usage was determined by monitoring cylinder usage while patients used each delivery method at home. RESULTS: We found no difference in mean PaO2 (90 ± SD 13 vs 91 ± SD 11 torr) or PaCO2 (44 ± SD 8 vs 42 ± SD 6 torr) between the two delivery methods during home use. The cylinders lasted nearly three times longer with pulse-dose usage than with continuous-flow O2 (86 ± 2 vs 30 ± 1 h). Patients did not express a preference for one device over the other. CONCLUSION: Pulse-dose O2 therapy provides oxygenation equivalent to continuous-flow O2 while using only one third the amount of O2. This may reduce O2 cost and delivery costs and provide therapy for more patients when supplies are limited.