The global absolute poverty rates of the World Bank demonstrate a continued decline of poverty in developing countries between 1983 and 2012. However, the methodology applied to derive these results has received extensive criticism by scholars for requiring the application of PPP exchange rates and CPIs that are not constructed to capture the consumption habits of those who live in absolute poverty. Those methodological concerns cast reasonable doubts on the poverty rates reported. First, in this paper, I demonstrate the validity of the hypothesis that the World Bank’s method inconsistently measures global absolute poverty. Second, I introduce new estimates of global absolute poverty based on a consistent methodology suitable for comparisons in time and between countries. For this purpose, I follow a well known concept of measuring bare bones subsistence using a consumption basket. This absolute poverty yardstick tracks bare bones survival costs and is priced in domestic nominal terms. The minimum caloric requirements are calculated separately for each country and year based on the demographic composition. The exact composition of the baskets is determined separately for each combination of country and year. The non-food component contains, among others, clothing and fuel consumption for basic heating, linked to monthly average temperature data. The results validate the critique on the World Bank’s methodology. They demonstrate large discrepancies in levels, which I find in many cases several times lower of what they report. This difference is far from being a linear change in all countries, which in turn fundamentally changes the geography and development of global absolute poverty. A sharp post 1990 increase together with a thereafter modest but longer decline brings the 2012 estimate only 1% lower than 1990.