Abstract A review is given of the wide variety of predictions that results from a Landau-type of description of the nematic-isotropic phase transition. This includes a discussion of the nature of the order parameter and of the various types of possible phases, of the influence of external fields, and of the effect of inclusion of spatial variations of the order parameter. The various predictions are compared with the available experimental results. It is concluded that there is still no clear picture about the nature of the singularity near the nematic-isotropic phase transition. Though the assumption of classical (mean-field) critical behaviour seems to be incorrect, there is no conclusive proof which alternative applies.