Measurements of the tritium and deuterium content of atmospheric hydrogen are reported. The measured tritium is corrected for local contamination by industrial hydrogen. The samples cover the period from 1960 to 1965. The tritium content shows a marked increase in 1961 but rises only slowly during 1962 and 1963. There is evidence for a further increase in the tritium content starting in fall 1964. The explanation suggested is, that tritiated hydrogen is injected into the high stratosphere by thermonuclear explosions and mixed down into the troposphere with a 2—3 years delay. Thus one would ascribe the first increase to 1958 testing, the second to the 1961/1962 thermonuclear test series.