As we saw in chapter 1, the classical nucleation theory (CNT) qualitatively described physical phenomena but fails by several orders of magnitude to give a quantitative description (Fig. 1.17) [47, 50-53]. Glasses are materials particularly interesting to test theoretical predictions because nucleation and crystal growth are slow processes in these systems due to their high viscos- ity (typically 106-1010 Pa.s) and they can easily be studied after quenching. Faced with the significant limitations of the CNT, it seemed essential to pro- pose other theories that report the nucleation process better. The purpose of this chapter is to present alternative theories or developments of the CNT that have been proposed. We do not intend to present an exhaustive list but to give the main ideas, in particular the most recent and most appropriate theories for a solid/liquid transition. Due to the large number of different approaches and, sometimes, to their complexity, a complete description of all theories is irrelevant and this chapter will endeavour to explain the con- cepts involved in the different approaches, the reader can then deepen his knowledge with the bibliographical references. Emphasis will be given to two theories currently subject to a growing success and extending the Gibbs’s concepts beyond the CNT: generalized Gibbs’s approach and the two-stage model. The on-going development of the experimental techniques brings support for some of these theories and examples of experimental observa- tions will also be presented.