An apparatus to measure the outgassing rate of small parts in an ultrahigh vacuum environment was constructed using aluminum alloys. In order to test the reliability of the apparatus, we prepared test pieces and test chambers with various surface areas but with the same surface treatments, and measured the dependence of the outgassing rate on the surface area. The outgassing rate was measured using the throughput method. Before and during baking, the outgassing rate was proportional to the surface area, but after baking, the outgassing from a Bayard–Alpert (BA) gauge modified the proportionality. Residual gas analysis using a quadrupole mass spectrometer (QMS), which was specially prepared for use in extreme‐high vacuum [H. Ishimaru, J. Vac. Sci. Technol. A7, 2439 (1989)], showed that the main component of the outgassing before baking was H2O. After baking, H2was dominant if the outgassing from the BA gauge was subtracted. The outgassing from the BA gauge was composed of mainly H2and CO. The outgassing rate of the gauge was the dominant contribution to the background outgassing rate after baking and was about 6×10−10‐2×10−9Torrl/s. Moreover, the outgassing rate of the gauge depended on the ambient pressure. If the outgassing rate of the gauge is correctly subtracted from the total outgassing rate, the correct outgassing rate of the sample can be determined.