Pneumatic tube transport has been reported to aggravate the error in partial pressure of oxygen (PO(2)) measurements caused by air bubbles. The aim of this study was to clarify the effect of manual and pneumatic tube methods of sample transportation and different amounts of air bubbles on arterial blood gas analysis. Blood gas samples from 15 patients and a pooled wasted blood mixture with 3 different levels of PO(2) were analyzed to determine the effects of air bubbles and manual versus pneumatic tube transportation on PO(2) levels. PO(2) increased significantly in samples containing 10% air bubbles and was exaggerated by pneumatic tube transport (from 115.63 +/- 9.31 mm Hg to 180.51 +/- 11.29 mm Hg, p < 0.001). In samples with low PO(2) ( approximately 30 mm Hg), the measurement was not aberrant regardless of the method of transportation or the amount of air bubbles contained in the specimen. However, in samples with medium and high PO(2) (> 70 mm Hg), aberrances in measurements were noted even with only 0.5% air bubbles and regardless of whether the sample was transported by manual methods or pressurized tube. The increments of PO(2) correlated positively with the amount of air introduced into the specimens. Thus, the measured PO(2) increased 8.13 and 31.77 mm Hg when 0.5% and 10% air bubbles were introduced, respectively, to samples with medium PO(2) (p < 0.05). The interaction between the amount of air bubbles and the method of transportation was significant (p < 0.001). Trapped air in the syringe should be expelled as thoroughly as possible, since the presence of only 1% air bubbles can result in aberrance in PO(2) measurement. Samples for blood gas analysis should be carried in ambient pressure to the laboratory because pneumatic tube delivery systems significantly aggravate the air bubble-related aberrance in PO(2) measurement.