In order to reveal the effect of oxidation on thin blade walls, a new machine allowing tests up to 1250degreesC under controlled atmosphere has been designed. Creep tests were performed on MC2 single crystal superalloy at 1150degreesC, under hydrogenated argon and dry air, but also with a switch from one atmosphere to the other after reaching the steady state creep stage. The results point out the decrease of the minimum creep rate in case of tests performed or at least started under hydrogenated argon, compared with the value obtained under synthetic dry air. This effect of oxidation was attributed to the protective oxide scale formed under hydrogenated argon. The low growth rate of alumina layer leads to a thinner zone affected by metal consumption, which is assumed to be non bearing, and prevents from vacancy flux toward the alloy. The second point results in slowing down creep mechanisms controlled by diffusion and therefore dislocation motion and microstructure evolution. Thermogravimetric tests confirm the difference of oxidation kinetics regards to environment (hydrogenated argon and dry air). However, oxide scales have different microstructures on thermogravimetric and creep samples when tested under air.