Official data from the Mahatma Gandhi National Rural Employment Guarantee Act website is examined alongside independently collected survey and ethnographic data. The validity of the official data is verifi ed by comparing it to field-level observations. This is done at various levels, and in doing so, it is shown that in certain ways the official mgnrega data is robust. In other very important ways the data is shown to be highly problematic for these particular villages in Tamil Nadu, for which the exercise is done.