The hidden economy suddenly appeared in economic research in 1977, and has since attracted much attention from economists. The central problems are how to assess the scale and causes of this illicit phenomenon. Many different methods of measurement have been applied to different Western countries. Unfortunately these methods yield extremely different size estimates. They also appear to measure different things. This paper reviews the methods used and makes a strong case for direct methods. An example of a direct method used in the Netherlands is presented. The results are used to analyse the relation between hidden and regular activities.