Accidental discharges of the hazardous nuclear fission products 137Cs+ and 90Sr2+ into the environment, such as during the Fukushima Dai-ichi nuclear accident, have occurred repeatedly throughout the ‘nuclear age.’ Numerous studies of the fate and transport of 137Cs+ and 90Sr2+ in soils and sediments have demonstrated their strong and selective binding to phyllosilicate clay minerals, primarily by means of cation exchange into interlayer sites. The locally concentrated amounts of these radioactive beta-emitters that can be found in these host minerals raise important questions regarding the long-term interplay and durability of radioisotope–clay associations, which is not well known. The present study goes beyond the usual short-term focus to address the permanence of radioisotope retention in clay minerals, by developing a general theoretical understanding of their resistance to the creation of defects. The present study reports ab initio molecular dynamics (AIMD) calculations of the threshold displacement energy (TDE) of each symmetry-unique atomic species comprising the unit cell of model vermiculite. The TDE values determined are material specific, radiation independent, and can be used to estimate the probability of Frenkel-pair creation by direct electron–ion collision, as could be induced by the passage of a high-energy electron emitted during the beta-decay of 137Cs, 90Sr, and daughter 90Y. For 137Cs and 90Sr, the calculated probability is ~36%, while for 90Y the probability is much greater at ~89%. The long-term retention picture that emerges is that decay will progressively alter the clay interlayer structure and charge, probably leading to delamination of the clay, and re-release of residual parent isotopes. Further work examining the effect of Frenkel defect accumulation on the binding energy of parent and daughter radionuclides in the interlayer is thus justified and potentially important for accurate long-term forecasting of radionuclide transport in the environment.