Residual stress is a complex subject which requires clear understanding of the phenomenon before any serious attempt in measuring it can be made. A variety of different phenomena is often lumped together under one collective name as residual stress and to complicate it further, these various types of residual stress can exist simultaneously together in an elastic body. Hence when interest is focussed on a specific type of residual stress, not only a distinction in identifying the type of residual stress is required, but it is imperative to be able to separate the various types of residual stress during its measurement. The magnitude and distribution of residual welding stresses produced in plates due to fusion welding are determined from X-ray diffraction measurements. In the investigation, the gas metal-arc welding process, using CO₂ as shielding gas and 1.2 mm consumable solid wire as electrode, is employed. Residual stresses arising out of two different methods of metal transfer by this process, namely the spray and short-circuiting arc methods, are measured and compared. Effects of edge preparation and weld length upon the residual welding stress patterns are investigated. Results obtained so far from X-ray measurements have indicated large scatter of experimental points on the residual stress patterns. The author believes it can be attributed to local variation in initial stresses present in the plates before welding. Large variation in initial stress distribution can plausibly obscure the residual welding stress patterns sought in the present investigation. An interesting question immediately arises from this, namely how the initial stresses add on to the residual welding stresses during welding. This problem has been discussed in length in section 1.2. The main school of throught supports simple additiveness of the two types of residual stress. However, conclusive evidence is lacking in this conjecture. This is mainly attributable to the lack of 'the' method of residual stress measurement upon which reliability can be confidently placed. Various methods of residual stress measurement have been discussed in section 1.5 and the difference in characteristics is discussed in section 2.4. It would appear that each method has its own drawbacks which are serious when accurate measurements of residual stress are required. The main drawback with the X-ray methods is the shallow effective penetration of the diffracted x-rays, and thus the method could only reveal surface stresses present in the surface layer of the elastic body. Problems peculiar to the X-ray methods are discussed in Chapter Six where the effects of surface preparation and preferred orientation on X-ray stress measurements are presented.