Purpose - The contribution of this study is both methodological and empirical. It provides a method of estimating preference consistent true cost of living indices and demonstrates the use of unit values (food items), adjusted for quality and demographic effects, as prices. Using NSS data, changes in living standards (measured by per capita real expenditure) in India are examined between 1999/2000 and 2009/2010. The paper aims to discuss these issues. Design/methodology/approach - From the adjusted unit values, exact price indices are computed using QAIDS-based preference consistent methods that allow between-item substitution effects and variation across states. Findings - A comparison of the nominal and price deflated real expenditures under alternative temporal price scenario during 1999/2000-2009/2010 shows that the states largely preserve their ranks over the periods, in spite of differential temporal price movement. However, comparison of the nominal and price-deflated real expenditure growth reveals that the rankings are sensitive to the price deflator used. Practical implications - The results question the wisdom of the treatment of large countries with heterogeneous preferences, e.g. India, as single entities in PPP calculations as in the ICP project. Hence, the results have methodological and empirical implications that extend beyond India. Originality/value - The study provides evidence on the issue of spatial difference in the temporal movement in prices, where no such evidence exists, and contains the first evidence on living standards in India in the post global financial crisis era. Also, this is the first attempt to base calculation of temporal movement in prices, as measured by the exact price indices, on the adjusted unit values of food items.