A model for the effects of low applied stress on grain boundary segregation/dilution of solute has been suggested in the present paper. This model is based on the following assumptions: (1) The grain boundary is a weaker region on strength than the perfect crystalline in the interior of gain and will preferentially be deformed when a polycrystalline is exerted by an low applied stress. (2) Grain boundaries will work as sources of vacancies to emit vacancies when a compression stress is exerted on them and as sinks to absorb vacancies when a tension stress is exerted; (3) Oversaturated vacancies induced by the applied stress will be combined with the solute atoms to form vacancy-solute atom complexes, the diffusion rate of which is far greater than that of solute atoms in matrix; (4) The effects of applied stress on grain boundary segregation/dilution of solute will be controlled by the balance between the complex diffusion and the reverse solute atom diffusion. According to this model, there will be a critical time during stress aging, at which a maximum level of grain-boundary segregation/dilution will occur. This model can be corroborated by Shinoda and Nakamura's observation for phosphorus and Misra's observation for sulfur in steels. It can be expected that a new basis for understanding the low ductility intergranular fracture induced by applied stress will result from this new model.