Using quantile regression methods, this paper analyses the gender wage gap across the wage distribution and over time (1990-2014), while controlling for changing sample selection into full-time employment. Our findings show that the selection-corrected gender wage gap is much larger than the one observed in the data, which is mainly due to large positive selection of women into fulltime employment. However, we show that selection-corrected wages of male and female workers at the lower half of the distribution have moderately converged over time. The reason for this development have been changes in the composition of the male full-time employment force over time, which in spite of the rather constant male full-time employment rate, have given place to a small but rising selection bias in male observed wages. In the upper half of the wage distribution, however, neither the observed nor the selection-corrected gender wage gap has narrowed over time.