The article focuses on the development of the labour market in the Czech Republic from the perspective of employment and unemployment between 1998 and 2004. Using data from the Czech Statistical Office, and within a reference framework of EU countries, the authors discuss and analyse the factors that determine unemployment and employment in Czech society. The authors use latent logistic regression to verify the assumption that the odds of unemployment are not evenly distributed across the entire Czech population and to identify three segments of the labour market in the Czech Republic. In each of these three segments the employment and unemployment odds differ, and the factors that determine these odds function differently in each segment.