Cognitive interviews were conducted in the Washington, DC metropolitian area with owners, renters, households with infants, and large households with difficult compositions of relatives and nonrelatives to test changes in question wording in the tenure, age, and relationship questions. The results showed that eliminating "cash" from "cash rent" and "occupied without payment of cash rent" in the tenure question did not have a negative effect on respondents' understanding. Reversing the order of the age and date of birth questions work well, as did adding an instruction to "please report babies as age 0 when the child is less than 1 year old." However, the font and layout of the age question de-emphasized the reference date, and the ages of several people whose birthdays occured between the reference date and the interview date were misreported. Changing "nature born son or daughter" to "biological son or daughter" and "foster child" to "foster child or foster adult" worked well, as did changing the layout of the relationship categories to save space on the form. Recommendations concerning the font and layout of the age and date of birth questions are being included in the 2005 National Census Test.