Although polyamide "nylon 6" polymer is an important industrial material, there remain many questions about the details of the various structures and the conversion between them. Using the MSXX force field (developed previously from ab initio quantum calculations), we predict the crystal structures, folds, and lamellae of nylon 6, leading to the following results. (a) Assuming infinite chains and evaluating the free energy of all 112 regular crystal structures, we find three classes of crystal structures: alpha form, gamma form, and delta form. We find that at 300 K the CL form is most stable, with gamma and delta higher by 0.4 and 0.3 (kcal/mol)/(amide unit), respectively. We calculate Young's modulus in the chain direction to be 295 GPa for a., 135 GPa for gamma and 253 GPa for delta. These values are above the experimental value of 168 GPa for a form because the experimental system has a finite lamella thickness, disorder in the chain conformation, and imperfections in the crystallinity. (b) We find the thermostability of alpha form over other forms arises from intra-H-bonds in the a form, which are dynamically and entropically favored. (c) We propose five detailed steps in the transition between the alpha and gamma forms. We also identify the structures of the other two experimentally observed metastable forms, and delta. Our structures explain the available fiber X-ray results. (d),The H-bond schemes for all. regular crystal structures are examined. We find that the gamma form has a more linear (stronger) H-bond than alpha form, which is consistent with the interpretation from solid-state NMR. (e) Considering that nylon forms lamellae with finite thickness in the chain direction we considered all five possible loop structures and the two best (of eight) possible stacking schemes for the folded sheets together with the 14 possible sheet displacements. We find that the optimum lamella for a form has the alkane loop fold (one amide per loop) and packs so that adjacent sheets are displaced by +/- 3.7 Angstrom (3b/14), which is in good agreement with the conclusion from fiber X-ray. Our amide pocket model explains the observed sheet displacements in nylon 6 and nylon 66 and also the progressive shear in nylon 66 and nylon 46.