Semiparametric estimation has gained significant attention in the study of wage inequality between men and women in recent years. By extending the wage gap at the mean towards the entire wage distribution using quantile regression, it enables researchers to ascertain the direction and the proportions of differences in characteristics and returns to these characteristics at different parts of the wage distribution. This line of research has been prominent in western society but has not yet been explored in the context of the Malaysian labor market. To fill the gap, this paper examines the gender earnings gap in Malaysia between 1994 and 2004 using Malaysia Population and Family Survey data. The gender earnings differential, as measured by the log percentage point is 53% in 1994. The difference reduces to 45% for a restricted sample and 42% for the unrestricted sample in 2004. However, it was found that the gender wage gap reduces as we move up the wage distribution. This suggests that women suffer from a sticky floor effect, i.e., the gender wage gap is bigger at the bottom of distribution. More importantly, the observed gender wage differentials do not reflect differences in the productive characteristics of the workers. In fact, it accounts for very little, if any, of the gap in Malaysia. However, the extent of the price effect is larger at the bottom end of the distribution than at the top.