Abstract: Whether the long-term buoyancy in saturated clay is less than that in pure water is still a controversial issue. The theoretical analysis of this issue is lacked and the existing experimental researches lead to inconsistent opinions. A model test of up-lifting is conducted in both the cohesionless and cohesive saturated soils, and the reduction coefficient of buoyancy is obtained during the ultimate limit state of up-lifting (ULSU, i.e., the effective stress approaches zero). The experiment is based on geometry measurements, thus the inaccuracy of force measurement is avoid. The results indicate that the reduction coefficient of buoyancy is close to unity for different saturated soils. The uplift force in the saturated soil is approximately the same as that in pure water, and no significant reduction is observed. For the saturated soils, theoretical analyses based on the effective stress principle show that the reduction coefficient of buoyancy during the ULSU is the reciprocal of the Skempton’s B value before consolidation. However, the B-value significantly greater than unity is not observed for the varied saturated clays. Therefore, during the ULSU in the saturated soils, significant reduction in buoyancy is not supported by the experimental and theoretical results.