Statistical analyses have been undertaken using both the maximum likelihood method and Bayes paradigm on Charpy impact energy data obtained from C-Mn submerged arc weld metal specimens tested over a range of temperatures. These specimens were exposed for defined periods to a neutron spectrum containing a low ratio of thermal to fast neutrons under surveillance conditions at Magnox reactors in the UK. A subset of specimens was then exposed to a neutron spectrum of a high ratio of thermal to fast neutrons (Halden reactor, Norway). Thus the data include results obtained, whereby specimens previously exposed to irradiation in Magnox reactors under surveillance conditions were subsequently exposed to this high thermal to fast neutron dose spectrum. The result is that the upward shift in the ductile to brittle transition temperature arising from thermal neutrons saturates after a period of exposure. The statistical analyses address the contribution from neutron dose and rate to the changes in these data to evaluate the underlying contribution made by the thermal neutrons as compared with the fast neutrons. The results of the analyses are discussed with respect to the role of neutron dose and dose rate together with known metallurgical variables and the sensitivity to these parameters.