With a larger urban population than the United States of America in 2012, India presents an urban organisation that is becoming more complex. A previous study showed us the nationally stabilized and the regionally unbalanced situation of the Indian urban system. These Indian urban subsystems show signs of urban primacy or urban macrocephaly (Querci J., 2011). Thus, this situation highlights the complexity of the multiscalar organisation within the system and gives its complex nature to the system. This systemic organization involves, over time, an evolution of the urban logics causing an evolution of towns and cities dynamics. In response, the interactions between towns and cities transform the urban logics of the Indian urban system and its multiscalar organisation. The main issue of that research will be to highlight the evolution of the Indian urban system, considered as an ancient urban system having known a shock reorganizing its urban framework (Bretagnolle A. et al., 2009). Taking into account the evolution of the urban logics, this research will be conducted by addressing the evolution of the multiscalar organization of the Indian urban system since the early 20th century. The Indian census usually represents the Indian urban population within six classes of towns and cities. The intervals used keep the same construction rules (Chapman P. G., Pathak P., 1997). Nonetheless, if we want to analyze the evolution of the interactions between towns and cities, forcing the multiscalar organization of the system, we need to reconsider them. Thus, the Indian administration keeps the same cutting since the earlier census and we can question ourselves on their actual validity. As the urban logics evolve, we need to determinate size classes of towns and cities according to the size distributions of towns for each census year. The main aim is to find size classes of towns and cities able to emphasize the multiscalar dynamics occurring in the organization of the urban system over time. It is in that sense that the administrative size classes of towns and cities cannot be used if they remain the same over time. A. S. Garmestani et al.(2008) worked on this particular issue within the framework of an analysis on the evolution of the U.S urban system. To the construction of the size classes of towns, they developed a methodology basing on breaks within the rank-size distribution of American towns for a given time. In our case, the main idea of this methodology can be used and we will precise its particular development. We applied this methodology to the Indian rank-size distribution for each decade from 1901 to 2001. The outcomes allow us to highlight the evolution of the multiscalar organization of the Indian urban system over time. They also allow us to emphasize the stability of the Indian urban system over time as well as progressives changes of the urban logics occurring in its evolution.