The article aims to ascertain how different assumptions about population segmentation affect the estimated size of the shadow economy, measured according to a method proposed for the first time by Pissarides and Weber [1989]. The method uses household survey data and applies a set of criteria to divide the population into households that operate fully legally in the formal, registered segment of the economy and those that generate incomes in the unregistered, tax-evading segment of the economy, referred to as the shadow economy. The author measures the size of the shadow economy by modeling the relationship between income and expenses. He looks at how measurements can be affected by the use of different definitions of a specific set of criteria and also comes up with a set of completely different criteria for dividing the population than those usually proposed in the literature on the subject. A study based on U.S. data (Consumer Expenditure Survey) for the 1980-2003 period shows that the way in which the population is divided is of fundamental importance to estimating the size of the shadow economy with the Pissarides and Weber [1989] method, Dymarski says. Depending on the segmentation criterion, estimates varied significantly. The upper limit for the size of the shadow economy as a percentage of GDP ranged from under 10% to around 40%. However, findings by other authors suggest that the 40% level is improbable, according to Dymarski. The author also argues that the criterion on the basis of which the population is divided is far more important than the definitions used.