The paper presents the results of three full-scale tests that were carried out to investigate the influence of transient flooding on the performance of multi-anchor reinforced soil walls (MAW). The walls were constructed to a height of 6 m and flooded from the front of the structures to about mid-height and then drained in two to four stages. The walls were constructed with three different sand soils varying with respect to permeability, fines content and shear strength. Earth pressures and anchor loads were monitored during the flooding and draining stages and in-situ anchor pullout tests were performed. The analysis of results showed that the wall facings were very permeable and thus prevented unbalanced hydrostatic or seepage forces from developing during drawdown that could increase anchor loads beyond drained condition values. The accuracy of the current PWRC (2002) analytical model to estimate anchor capacity of MAW systems was demonstrated to vary widely depending on soil type and whether the soil was flooded or drained. However, on the load side, the method proposed by Miyata et al. (2009) to predict anchor loads was judged to be more accurate than the current BS8006 (2005) and PWRC (2002) methods based on the mean and spread of bias statistics where bias is defined as the ratio of measured to predicted anchor load.