All too often those making welding stress measurements fail to consider what length scales and types of residual stress are important from a materials or component performance viewpoint. Unwittingly this can lead to inappropriate measurements and interpretations. Traditionally neutron diffraction has been used to infer macro (type I)-residual stresses, but type II and type III residual stress information is also contained within the diffraction peak shifts and widths. In this paper we examine the residual stresses generated at different scale levels within an austenitic stainless steel welded plate containing weld stop-start features such as those typically encountered in industry. Significant increases in the macro-stresses are found at these locations. In addition, calibration samples that were free of macro-stress but had undergone known amounts of plastic strain were also measured, allowing diffraction peak broadening measurements to be discussed in terms of plastic strain (type III stress). Finally, an analysis of the individual diffraction peaks allowed the level of type II intergranular stresses to be determined and thereby the extent of plastic strain inferred and related to the welding process. This stresses can be important as weld metal itself is anisotropic in nature and weld measurements based on single diffraction peak measurements can be significantly affected by type II intergranular stresses, leading to stress levels deviating significantly from the underlying macrostress.