The Zipf’s law is studied here in the context of size distribution of Indian cities and the power law exponent is estimated. We have used the data from the Indian censuses of 1981,1991 and 2001. The analysis shows that the population distribution in Indian cities do follow a power law similar to the ones found in other countries. The scaling exponent are found to be 2.15 ± 0.01 for 1981, 2.11 ± 0.01 for 1991 and 2.05 ± 0.02 for 2001 from the linear fit. We have also estimated the scaling exponent from the maximum likelihood estimator technique which is found to be 2.04 ±.07 for the year 2001.