Estimates of the uncertainty ΔD in predicting the achievable tritium breeding ratio (TBR) due to the uncertainties in nuclear data base are presented for several fusion blanket concepts. Specifically, the impact of the current uncertainties in measuring basic nuclear data on the calculated TBR is analyzed and discussed for four leading blanket designs that utilize different breeding materials, namely, Li2O, 17Li-83Pb, LiAlO2, and Flibe. The impact on the TBR values of various evaluations for beryllium, which is employed as a multiplier in the latter two blankets, has been studied. Estimates for ΔD in other blanket concepts have also been assessed. Moreover, estimates have been made, based on previous studies, for the contribution to ΔD introduced by using neutron cross-section libraries that have different group structure and weighting spectra. Based on statistically incorporating the present cross-section uncertainties and their correlation in the analysis, the range of the uncertainty in TBR was found to be between 2 and 6% in all the concepts considered. The nonstatistical treatment for cross-section errors tends to give larger values for ΔD. The uncertainty in TBR introduced by misrepresenting the secondary energy-angle distribution of the 9Be(n,2n′) cross section ranges from ∼4% in the Flibe to ∼2% in the LiAlO2 blanket. Uncertainty up to ∼15% can be encountered in the TBR evaluation in thin blankets with natural 6Li enrichment if broad-group cross-section libraries are used. However, this uncertainty can be reduced upon using an appropriate weighting spectrum representative of the one found in these blankets type.