The aim of this study is to examine whether, and to what extent, the differences in wages between women and men can be explained by discrimination against women in the labour market.The first part of the analysis is a cross-sectional analysis. Firstly, a model is estimated with wages as the dependent variable and with sex, individuals' qualifications, personal characteristics, occupation and the branch of industry in which they are employed as the independent variables. The results of this regression reveal that the hourly wage for a woman is about nine SEK lower than for a man. Separate wage functions for women and men show that women obtain a lower return from both education and work experience. A breakdown of the difference in wages reveals that most of the dissimilarity is explained by discrimination against women and very little by differences in men's and women's productivity. In the absence of discrimination, women's relative pay would be between 16 and 25 percent higher. The stability of the findings is checked by means of reverse regression. This resulted in discrimination against men in some cases and against women in others. Thus, different conclusions can be drawn from the two methods. In certain special cases the results produced by reverse regression are correct, whereas, in more general cases, the direct method shows itself to be more satisfactory. The study ends with an analysis of the effects of various laws and agreements on the development of women's wages and employment in Swedish industry. The introduction of equal pay, the removal of the ban on night work for women in industry and the wage solidarity policy have had a positive influence on the relative demand for women workers. The study's conclusions are, firstly, that the differences in pay between the sexes can partly be explained by discrimination against women. Secondly, within industry, wage discrimination against women has declined as a result of changes leading in an anti-discrimination direction. / digitalisering@umu