Nucleation and motion of kink pairs on partial dislocations are examined by elasticity theory for materials with a high Peierls stress. Two approaches are used: one considers the change in average stacking-fault energy due to alloying elements and the other considers the change in local SFE due to a nearby solute atom. Both approaches highlight the role of SFE on kink nucleation, propagation and annihilation and both furnish strain rate as a function of stress, temperature and SFE. Model predictions are (...) compared with yield stress data for two systems: firstly, an intermetallic, MoSi 2 , which softens for alloying elements that decrease the SFE and hardens for Re additions that increase the SFE; secondly, a ceramic oxide, MgO-Al 2 O 3 spinel, which softens with increasing addition of excess alumina and at the same time exhibits a decrease in SFE. The average SFE approach agrees qualitatively with the data while quantitative agreement is obtained with the local SFE approach. The possibility is considered that the model applies to other materials, such as TiAl, HfV 2 and Fe 3 Al, which show softening with certain alloying additions. One requirement is that the dislocations are dissociated more than a few atomic distances; otherwise kink nucleation occurs on the perfect dislocation . Hence the model does not apply to materials such as bcc metals which only have a core dissociation. Normal hardening effects of solutes from size and modulus misfits are additive with any softening effects from a decrease in SFE and so may mask the latter, as occurs for W additions to MoSi 2. (shrink)