As long as it is employed cautiously enough, the model approach is a useful tool to estimate simultaneously the size and the development of the shadow economy in several countries. However, a second method is necessary to calibrate the model. The currency demand approach can lead to highly implausible results; the size of the shadow economy might be largely overestimated. An alternative is the survey method. For real tests of whether a variable has an impact, procedures are necessary that do not use the same variables as those used to construct the indicator. Thus, to make progress in analysing the shadow economy, the model approach has a role to play, but it has to be complemented by other methods employing different data. The currency demand approach cannot be used as long as it employs the same variables for its constructions.