A new methodology is developed for the prediction of RPV embrittlement that utilizes a combination of domain models and nonlinear estimators including neural networks and nearest neighbor regressions. The Power Reactor Embrittlement Database is used in this study. The results from newly developed nearest neighbor projective fuser indicate that the combined embrittlement predictor achieved about 67.3 % and 52.4 % reductions in the uncertainties for General Electric Boiling Water Reactor plate and weld data compared to Regulatory Guide 1.99, Revision 2, respectively. The implications of irradiation temperature effects to the development of radiation embrittlement models are then discussed. A new methodology that incorporates the chemical compositions into the Charpy trend curve is also developed. The purpose of this new fitting procedure is to generate a new multi-space topography that can properly reflect the inhomogeneity of the surveillance materials and utilize this multi-space trend surface to link and project the surveillance test results to that of reactor pressure vessel steels.