Results of the self-cooled, liquid-metal blanket design from the Blanket Comparison and Selection Study (BCSS) are summarized. The objectives of the BCSS project are to (a) define a small number (about three) of blanket concepts that should be the focus of the blanket research and development (R&D) program, (b) identify and prioritize the critical issues for the leading blanket concepts, and (c) provide technical input necessary to develop a blanket R&D program plan. Two liquid metals [lithium and lithium-lead (17Li-83Pb)] and three structural materials [primary candidate alloy (PCA), ferritic steel (FS) (HT-9), and vanadium alloy (V-15 Cr-5 Ti)] are included in the evaluations for both tokamaks and tandem mirror reactors (TMRs). There are major differences in relevant design parameters between a tokamak and a TMR, such as surface heat flux, first-wall erosion rate, and magnetic flux density. As a result, the magneto-hydrodynamic (MHD), heat transfer, and structural requirements for a tokamak reactor are much more stringent than that of a TMR. This has a significant impact on the design philosophy for the blankets. The reference design for the tokamak reactor is the poloidal/toroidal flow module, whereas that for a TMR is of the tube configuration similar to the Mirror Advanced Reactor Study design. Analyses were performed in the following generic areas for each blanket concept: MHD, thermal hydraulics, stress, neutronics, and tritium recovery. Integral analyses were performed to determine the design window for each blanket design. The Li/Li/V blanket for tokamak and the Li/Li/V, LiPb/LiPb/V, and Li7Li/HT-9 blankets for the TMR are judged to be top-rated concepts. In general, the blanket concept of a TMR is ranked higher than that of a tokamak reactor for the same coolant/structural material combination. This is the result of less stringent design requirements for a TMR compared to that of a tokamak reactor. Because of its better thermophysical properties and more uniform nuclear heating profile, liquid lithium is a better coolant than liquid 17Li-83Pb. From an engineering point of view, vanadium alloy is a better structural material than either FS or PCA since the former has both a higher allowable structural temperature and a higher allowable coolant/structure interface temperature than the latter. Critical feasibility issues and design constraints for the self-cooled, liquid-metal blanket concepts are identified and discussed.