Economic output implies that underground sectors such as prostitution are taken into account. This article presents an innovative methodology to measure turnover and added value in prostitution based on a combination of observational and Internet data. The method is applied to Belgium. Turnover is broken down in transactions and price per segment. The starting point is an observation-based measure of turnover in one locational and visible segment of the market: window prostitution. Fundamental differences between segments make linear generalizations from one segment invalid. Therefore, we estimate the relative size of transactions in other segments (such as brothels or escort) with Internet data. In combination with measures of average price per transaction, a consolidated estimate of turnover in prostitution in Belgium is measured. Estimates of nonresident production are based on data on sex workers’ country of origin. Several bootstrap replications allow for robustness checks of the delta-based standard errors.