The standard cumulative normal distribution has been approximated by over 70 formulas. Very recently, Eidous and Abu-Shareefa (2019) have reviewed 45 competitive formulas given in literature from 1945 to 2019, to approximate the standard cumulative normal distribution. They introduced nine more accurate but quite complicated formulas in computation. However, they have recommended an approximation “FSE(z)" given by Soranzo and Epure (2014) on the basis of simplicity and the low maximum absolute error. This has motivated us to re-examine some competitive approximations and propose a new approximation based on fitting a cubic regression model, to fit the normal cumulative probabilities over some ranges of the standardized values.