The Part I of this study [Yang JL, Zhang Z, Schlarb AK, Friedrich K. Polymer 2006;47:2791-801] provided systematic experiments and General discussions on the creep resistance of polyamide 66 nanocomposites. To promote these works, here we present some modeling and prediction attempts in order to further understand the phenomena and mechanisms. Both a viscoelastic creep model named Burgers (or four-element model) and an empirical method called Findley power law are applied. The simulating results from both models agree quite well with the experimental data. An additional effort is conducted to understand the structure-property relationship based on the parameter analysis of the Burgers model, since the variations in the simulating parameters illustrate the influence of nanofillers on the creep performance of the bulk matrix. Moreover, the Eyring stress-activated process is taken into account by considering the activation volume. Furthermore, in order to predict the long-term behavior based on the short-term experimental data, both the Burgers and Findley models as well as the time-temperature superposition principle (TTSP) were employed. The predicting capability of these modeling approaches is compared and the Findley power law is preferred to be adopted. Master curves with extended time scale are constructed by applying TTSP to horizontally shift the short-time experimental data. The predicting results confirm the enhanced creep resistance of nanofillers even at extended long time scale. (c) 2006 Elsevier Ltd. All rights reserved.