Abstract This paper compiles alternative estimates of underground economies in twenty five transition countries during the transition decade and finds a disturbing lack of convergence between them, calling into question the reliability of GDP figures (which in varying degrees now include non-transparent imputations for the “non-observed economy”) as well as the macro model estimates of the unrecorded economy. A corollary of this finding is that substantive results from many studies examining the consequences of the radical transition from planned to market economies must be viewed with considerable skepticism. Underground (unobserved, non-observed, unrecorded) economic activities play a major role in transition economies. Evaluations of the success and failure of the transition experience should be based on estimates of total economic activity (TEA) namely, recorded plus unrecorded economic activity. We examine the conceptual and empirical relationships between new National Income and Product Accounts (NIPA) methods for obtaining “exhaustive” measures of total economic activity and the two most popular macro-model approaches (electric consumption and currency ratio models) for estimating the size and growth of the unrecorded sector. Our updated empirical results detailing the size and trajectory of unrecorded activities obtained from different estimation methods reveal a disturbing lack of convergence. Until these important differences are resolved, investigations of the relationship between economic reforms and economic outcomes during the transition decade must be viewed with considerable caution. Given the shortcomings of conventional macro model estimates of the underground economy and the lack of transparency and consistency of NOE estimates, it is high time that the profession acknowledges how little we really know about underground economies and their causes and consequences.