Material flow analysis (MFA) is a tool for research and decision support in environmental policy and management. In order to promote the use of MFA at different spatial scales, a quantification of the uncertainty in nationwide, regional, and urban MFA methodologies is provided. In particular, the impact of the input data quality on the main MFA indicators is analyzed and the sources and extent of uncertainties for different spatial scales are listed. The types, origin, and extent of the errors are described in detail and several imputation methods are explained and evaluated. By introducing a novel approach to account measurement errors in data sets with 'very few details on the measurement errors,' this article aims at contributing to the development of a standardized method to account for the uncertainty in MFA studies. This study uses the time series of MFA data for 1996-2011 at three spatial scales—nationwide (Sweden), regional (the Stockholm Region), and metropolitan (Stockholm, Gothenburg, and Malmo)—to determine how propagation of measurement errors affects the MFA results. The following MFA indicators were studied: direct material input; domestic processed output; and domestic material consumption. Generally, availability decreased as the spatial scale was lowered, whereas data errors increased. In the specific case of Sweden, the data on freight transport by rail and on waste produced by economic activities at the regional and metropolitan level should be improved.