Ti-48Al-2Cr-2Nb (at%) with a g phase structure was tensile-tested in a high purity hydrogen gas in a temperature range of room temperature to 973 K. The reduction of elongation was observed in hydrogen gas in the entire temperature range, although the fracture surface changed mainly from cleavage to grain boundary facets both in vacuum and in hydrogen gas as the temperature increases. From transmission electron microscopy, fractography, the content of hydrogen introduced during tensile testing, and consideration of the diffusion rate of hydrogen, the hydrogen environmental embrittlement of the TiAl based alloy is suggested to result not from the formation of hydrides but from local accumulation of solute hydrogen introduced by dislocations from the test environment during testing. The hydrogen introduced during tensile deformation mainly decreases not grain boundary strength but cleavage strength, and then the reduction of elongation in hydrogen gas is caused by the decrease of the transgranular strength in the wide temperature range, although the grain boundary strength may also decrease with increasing temperature. KEY WORDS: TiAl based alloy; hydrogen gas; environmental embrittlement; tensile property; elevated temperature.Al: 33.52 (48.2 at%), Nb: 4.71 (1.97 at%), Cr: 2.71 (2.02 at%), Si: 0.007, Mn: Ͻ0.005, Fe: 0.10, Ni: 0.02, Cu: Ͻ0.005, C: 0.008, O: 0.063, N: 0.0041, H: 0.0009, Ti: bal. The alloy ingot of about 20 kg was prepared by inductionskull-melting in an argon atmosphere, and then the alloy bars of 16 mm in diameter were prepared by hot extrusion at 1 473 K. The extrusion ratio was 7.2.Tensile specimens with a gage length of 15 mm and a gage section of 1.5ϫ4 mm were machined using an electric discharging machine (EDM). The tensile direction of the specimens was parallel to the longitudinal direction of the extruded bar. After mechanical polishing, they were heat treated at 1 313 K for 50 h in an argon atmosphere in order to obtain a g-single phase structure. After mechanical polishing followed by electro-polishing, the specimens were tensile-tested in a temperature range of room temperature to 973 K using a screw-driven testing machine at a constant crosshead speed of 0.2 mm/min (a nominal strain rate of 2ϫ10 Ϫ4 s
Ϫ1). The test environments were vacuum (Ͻ4ϫ10 Ϫ4 Pa) and a flowing high purity hydrogen gas (99.99999 %) of 1 atm. The test chamber, which was attached to the testing machine, was evacuated to better vacuum than 4ϫ10 Ϫ4 Pa, and then backfilled with the high purity hydrogen gas. This process was repeated four times, before the specimens were tested in the flowing hydrogen gas. The tensile tests were carried out after the specimens were heated to a given temperature followed by holding for 15 min in each environment. In most cases, three specimens were tested in each condition, and the average values were used for the tensile properties.The fracture surfaces and microstructures of the specimens were observed using scanning electron microscopy (SEM) and transmission electron microscopy (TEM)....