Numerous studies have been performed on casein micelles because they have characteristic structure and biological functions, and play important roles in processing of milk. More than twenty models have been proposed since Waugh proposed the first model of casein micelle in 1958. In this review, models of the casein micelle were divided into three groups of early stage, submicelle, and nanocluster modeles, and then their characteristics were described. Submicelle models of Slattery, Schmidt and Walstra had been accepted by many researchers. However, since Holt proposed the nanocluster model in 1992, most of the proposed models were modified nanocluster ones. We made discussions on the electron micrographs which played a key role in proposing the nanocluster models. Finally, we described whether it is possible to explain the changes in casein micelles during processing milk using the submicelle model of Schmidt, nanocluster models of Holt, Horne, and Dalglesish & Corredig. It is impossible to explain all phenomena which occur in casein micelles during processing of milk using any models. Further studies on casein micelles are needed.