Graphene-based fillers are promising candidates for improving the long-term dimensional stability of structural epoxy resins. The development of reliable accelerated test methods is critical to predicting their durability. Herein, the viscoelastic behaviour of the industrial-grade amine-epoxy and corresponding nanocomposites with modified graphene oxide (up to 1.72 wt%) is studied using dynamic mechanical thermal analysis (DMTA) in both temperature and frequency sweep modes. Time-temperature (TTSP), time-water ageing (TWSP), and coupled superposition principles (TTWSP) are used to predict the long-term storage moduli of nanocomposites. The temperature and water shift functions, determined by curve shifting, agree well with the Arrhenius calculations based on data from independent tests. A methodology for constructing master curves from DMTA temperature scans is developed, and the results are well correlated with the predictions made from traditional isothermal multifrequency test data. The temperature and time dependences of the storage moduli are well fitted by known relaxation models, and their parameters are analysed in terms of the filler and water-ageing effects. The results can be applied in the development of the material model and the prediction of macromechanical properties and the service lifetime, which are relevant to examined epoxy/GO composites and other water-sensitive glassy polymers.