This study addresses the need to improve the reliability of measurements of weld residual stress; particularly for the type of welds used within the nuclear power industry. Such measurements are essential for predicting stress corrosion cracking and similar vital structural integrity issues.The study also forms part of a larger project aimed at improving the reliablility of weld residual stress measurements in general. Within this larger study it is intended to examine issues relating to all aspects of such measurements, from fundamental limitations to experimental practice and best technique. The emphasis in this proposal is the problem of un-stressed lattice parameter determination. In particular we seek to evaluate a relatively new method which was concieved specifically to improve the measurement of the un-stressed lattice parameter in weld specimens and other complex samples.