In this research, a kind of 2-acrylamido-2-methylpropanesulfonic acid sodium salt- (AMPS-Na-) type copolymer additive, the fluid loss additive (FLA), named as FLA A additive, was used for research. The performance of FLA A was tested and found to fail in the effective control of free water and to hinder the hydration process for delaying the breaking of the early hydration shell. The reason for it was the absorbed behavior and chelating effect of the AMPS-Na unit to Ca ²⁺ hydrating cement particles. Thus, a cationic polyvinyl alcohol (PVA) polymer, modified by glyoxal and boric acid, was discovered due to its excellence in associating with the FLA A additive for controlling the free cement-based material water amount and preventing the chelating effect of FLA A chains on the surface of the cement-based material. Glyoxal/boric acid-modified polyvinyl alcohol, abbreviated as PVAGB or PVA-G-B, was with special molecular properties, i.e., positive ZETA ( ζ ) potential characteristics and cross-linked molecular structure. Due to competitive absorbed behavior of glyoxal-modified hydroxyl groups and free Ca ²⁺ released by the hydration product, the chelating effect of AMPS-Na units to Ca ²⁺ was weakened and the possibility of FLA A chains being absorbed to the surface of the cement-based material was decreased. Then, the formation of a complete fluid loss system was obtained; i.e., the fluid loss volume decreased to less than 50 mL at 30°C and 108 mL at 80°C with 0.2 percentage by weight of cement (%BWOC) of PVAGB and 0.50%BWOC (percentage by weight of cement) of FLA A. Besides, the hydration process of cement-based material was accelerated due to formation of more C-S-H gels in the early hydration period. As a result, the cement-based material not only showed no worse compressive-strength retrogression but also showed a stable 28-day compressive strength of 28 MPa.