Data needs in the area of disability are changing with the passage of the Americans with Disabilities Act and with welfare reform. In response to a request from experts in the area of disability, a set of questions designed to measure disability was developed by the OMB Task Force on Disability Questions for Census 2000. These questions used a different measurement approach than the questions designed by the Census Bureau and proposed for the 1998 Dress Rehearsal. This paper presents the results of a cognitive evaluation of both sets of questions. The results showed that there were problems with both sets. The problems with the OMB questions were more substantive while the problems with the Census questions were related to format. The paper contains a recommendation for the 1998 Dress Rehearsal that combines the advantages of both versions.