An attempt is made, through modeling, to account for the decline in the 14C/C ratio in atmospheric CO2 after its bomb-test-induced peak in 1963. The model suggests that as of 1964 about one third of the bomb 14C remained in the stratosphere and that it was released to the troposphere with an e-folding time of about seven years. By contrast, measurements carried out in the stratosphere suggest that at that time the excess was closer to one quarter of the total and that the e-folding time for its decline was 3¿1 years. The anomaly between model and observation cannot be attributed solely to an inadequacy in the representation of the terrestrial biosphere. Rather, it must reflect either an inadequacy in the ocean model or in the measured stratospheric inventories. ¿ American Geophysical Union 1994