During our attempt of the precise determination of trace gases in the atmosphere by using atmospheric krypton (Kr) as an internal standard, we noticed that the atmospheric Kr concentration of 1.14 +/- 0.01 ppmv which had been quoted through half a century was different from the observation. The Kr concentration we obtained with three step concentration followed by GC/ MS measurement was 1.099 +/- 0.009 ppmv, about 4% lower than the currently used value.