We are presenting material flow accounting and related indicators for an Indian adivasis village in 1983 (Sarowar, Dangs, Gujarat). It gives a point of comparison with modern nation-wide material flow accounting. The aim is to test the feasibility of indicators of dematerialization of the economy in poor economies. We measured the annual material flows within the Sarowar village (670 inhabitants) in 1982-1983. The method was a combination of surveys, real time measurements, indirect measurements and laboratory dry matter measurements. The results were translated into recent concepts of material flow accounting (MFA), and compared with nation-wide studies. The total material requirement (TMR) of Sarowar (excluding air and water), USA, Japan, Germany and The Netherlands is respectively about 5, 84, 46, 86 and 84 tons per capita per year. The input (all biotic materials are expressed in tons dry matter) totalised 15.8 t DM cap−1 y−1 in Sarowar, which consists mainly of air (11 t cap−1 y−1) and biotic primary materials (4.1 t DM cap−1 y−1). The latest was composed of 29% of pastures, 25% of branches for field burning, 35% of fuel wood, 6% fodder, 1% of construction wood and 4% of grains. The outputs (15.8 t cap−1 y−1) were dominated by CO2 (15.1 t cap−1 y−1). In contrast, the output of The Netherlands (66.8 t cap−1 y−1) is dominated by export with air emissions (19 t cap−1 y−1), export (16 t cap−1 y−1) and embedded export (29 t cap−1 y−1). The apparent eco-efficiency (kg per US dollar, excluding air and water, including hidden flows) is 70, 3, 3, 3 and 3 kg $−1 respectively for Sarowar, Japan, USA, Germany and The Netherlands. The corrected eco-efficiency using Purchasing Power Parity is less contrasted with respectively 18, 3, 3, 4 and 3 kg $−1. Traditional human ecosystem measurements can serve as a basic comparison point, and as a test for dematerialization indicators. The limit of the indicator of eco-efficiency resides in the different degrees of monetization of the economies. In less monetized economies, this indicator is highly biased by the underlying non-market material flows. We discuss the use of ratios of non-substitutable factors in dematerialization assessment and we suggest the use of multi-criteria analysis instead.