Accurate prediction on long-term settlements of soft soils is challenging. One of the reasons is the time-dependent soil behaviours. How to extrapolate such behaviours from thin laboratory testing with limited time scales to in situ thick soil layer with large time scales is still an ongoing debate, which can be dated back to Ladd et al. (1977). Many experimental results previously used to advocate Hypothesis A have been re-analysed and found to align with Hypothesis B. However, discrepancies remain in some historical data, particularly in the case of Osaka clay retrieved from the seabed under the Kansai International Airport Islands. This study begins by introducing the main implications and selected existing models for the time-dependent behaviours of soft soil. It then focuses on the re-examination of selected testing results from Watabe et al. (2008b), supplemented by numerical simulation using isotache models. The analysis emphasises the importance of considering equal initial conditions when comparing data from samples with varying thicknesses. Furthermore, the good agreement observed between measured data and the simulation results using Hypothesis B methods demonstrates the validity of Hypothesis B for predicting the long-term consolidation behaviour of soft soil in both laboratory and field scales.