The primary goal of the Blanket Comparison and Selection Study (BCSS) was to select a limited number of blanket concepts for fusion power reactors, to serve as the focus for the U.S. Department of Energy blanket research and development program. The concepts considered most seriously by the BCSS can be grouped for discussion purposes by coolant: liquid metals and alloys, pressurized water, helium, and nitrate salts. Concepts using pressurized water as the coolant are discussed. Water-cooled concepts using liquid breeders — lithium and 17Li-83Pb (LiPb) — have severe fundamental safety problems. The use of lithium and water in the blanket was considered unacceptable. Initial results of tests at Hanford Engineering Development Laboratory using steam injected into molten LiPb indicate that use of LiPb and water together in a blanket is a very serious concern from the safety standpoint. Key issues for water-cooled blankets with solid tritium breeders (Li2O, or a ternary oxide such as LiAlO2) were identified and examined: reliability against leaks, control of tritium permeation into the coolant, retention of breeder physical integrity, breeder temperature predictability, determination of allowable temperature limits for breeders, and 6Li burnup effects (for LiAlO2). The BCSS's final rankings and associated rationale for all water-cooled concepts are examined. Key issues and factors for tokamak and tandem mirror reactor versions of water-cooled solid breeder concepts are discussed. The reference design for the top-ranked concept—LiAlO2 breeder, ferritic steel structure, and beryllium neutron multiplier — is presented. Finally, some general conclusions for water-cooled blanket concepts are drawn based on the study's results.