It should be noted that the previous estimates of the achievable level of accuracy referred roughly to 1970, and the discrepancy in the estimates of the accuracy of the calculated predictions are associated first and foremost with the difference inthe estimates of the indeterminacies of the initial group cross sections. The spectrum of the sensitivity of keff and CB towards different forms of the neutron cross sections in the present paper and, e.g., in [4-6], as a comparison shows are identical. Thus, changes in the estimates of the accuracy of the calculated predictions for kef f and CB of a fast breeder-reactor are dependent upon the consideration of the extensive data obtained during the last decade in differential neutron-physics experiments. Moreover, the predicted values of the achievable accuracy of the neutron cross sections in many cases are of a qualitative nature and are based on approximate estimates of the accuracy of the cross sections according to certain "recommendations ," based on agreements between experts [12]. They do not take into account all the sources of errors in the differential measurements, and also the effect of a specific method of estimating the cross sections, on which the correlation properties of the errors in the group cross sections may depend. The estimate of the accuracy of the calculated predictions of kef f and CB obtained in this present paper is based on the latest data of differential neutron-physics experiments. The feasibility of this accuracy has emerged because of the development of a system of group cross sections BNAB-78 [13], based on which the system of cross sections BNAB-MIKRO [14] was proposed, compiled from new estimates of the results of differential neutron data measurements. In parallel with the estimates of the most probable values of these data, their errors were also determined, which were represented in the form of a eovariational matrix W of the group cross sections, i.e., in a form suitable for practical application. It is important to emphasize that the covariational matrix describes the accuracy of the group cross sections BNAB-MIKRO: the errors of the experiments themselves are taken into account in it, and it is on these errors that this system of cross sections has been compiled. The BNAB-78 system differs from the BNAB-MIKRO system by taking account of the results of macroexperiments . Its use stipulates a higher accuracy of the calculated results [14].