The U.S. Census Bureau aims to count each person living in the United States just once and only once, yet some people are duplicated or listed at more than one residence, for various reasons. Since Census 2000, computer matching is employed to identify potential duplicates. The Targeted Coverage Follow-up (TCFU) Interview has been developed to collect information about the living situations of the suspected duplicates. The information is to be used to resolve where those suspected duplicates should be counted, using the Census Residence Rule that determines people’s usual residence by where they lived or stayed most of the time or on Census Day (April 1st) if there was no usual residence. In a large-scale cognitive testing conducted in 2010 and 2011, the TCFU questionnaire was tested with actual 2010 Census participants, including 226 suspected duplicates or a proxy from their household. Duplication was revealed and coded for 116 cases for analysis in this paper. We examined the completeness of the address and dates in the TCFU to determine whether the Census Residence Rule can be appliedautomatically without clerical intervention. We also looked at the success of applying the Residence Rule by different types of duplications. The findings suggested that there is possibility for automation in resolving duplication as part of the TCFU administration, while clerical effort is likely warranted to resolve inconsistencies in confirmed duplicates where the respondents did not always provide complete information (i.e. the address itself and the dates lived there) about the duplicate address, or when they were not certain of their whereabouts on the Census Day.