This paper estimated the liquefaction potential of a saturated soil deposit subjected to a horizontal seismic excitation at its base using the total stress approach. A comparative analysis between the simplified and the nonlinear dynamic methods was used to verify to what extent the simplified method could be reliable. In order to generalise the reliability of the simplified method for any value of the maximum acceleration for the used earthquakes, a correction for the maximum acceleration less than 0.3g was proposed based on the comparison of safety factor values determined by the dynamic method illustrated by the equivalent linear model with lumped masses and the simplified method for a given profile of soil subjected to 38 earthquakes. The nonlinear behaviour of soil was represented by two hyperbolic models: Hardin and Drnevich, and Masing. To determine the cyclic resistance ratio (CRR), the cone penetration test (CPT) based method, the standard penetration test (SPT) based method, and the shear wave velocity based method were used. The safety factor was calculated as the ratio of CRR/CSR, where CSR represents the cyclic stress ratio. The results of the proposed correction have given smaller values of the safety factor compared to the nonlinear dynamic methods for the maximum acceleration less than 0.3g. In other words, by considering this correction, the most unfavourable case is always given by the modified simplified method.