We conducted the first randomized controlled study to assess the effects of short-term 28% and 100% oxygen on PaCO(2) and peak expiratory flow rate (PEFR) in patients with acute severe asthma. Seventy-four patients (mean age, 37.9 +/- 9.7 years [+/- SD]; PEFR, 41.0 +/- 12.1% of predicted) from two emergency departments were randomized to receive 28% or 100% oxygen during 20 min. The administration of 100% oxygen significantly increases PaCO(2) (p = 0.03) and decreases PEFR (p = 0.001) as compared with administration of 28% oxygen. PaCO(2) before and during oxygen administration correlated significantly (p = 0.001) in both groups. Patients breathing 28% oxygen experienced a PaCO(2) fall; on the contrary, patients who received 100% oxygen showed an increase in PaCO(2), particularly those with PaCO(2) before oxygen treatment > 40 mm Hg. This study confirmed previous observations that oxygen dose should be variable and based on achieving and maintaining target arterial oxygen saturation measured by pulse oximetry > or = 92% rather than on prescribing predetermined concentrations or flow rates of inspired oxygen.