It is well known that in the unresolved resonance region, cross sections are treated by a statistical formulation and that the nonuniqueness of the mean resonance data set used to represent the cross sections for various fertile and fissile nuclides has an impact on the calculations of the Doppler coefficient and the fission and capture rate distributions of the heavy elements in a fast power reactor. In this Note, the corresponding uncertainty associated with the calculation of the central Doppler worth in the critical assemblies is indicated. Through the choice of the mean resonance data sets, it is possible to vary the self-shielded cross sections to some extent and to vary the temperature derivatives of the self-shielded cross sections by a significant amount so as to improve the agreement between experiments and calculations for central Doppler worths. This adjustment can be performed without significantly altering the infinite dilution cross sections that are recommended in the unresolved resonance region. This approach is illustrated with sample calculations performed for the central 235U Doppler worth in the ZPR-6-7 assembly.