The 1992 Simplified Questionnaire Test, (SQT) was developed to evaluate alternative short form designs in an effort to improve response rates in future censuses. Improvement to response rates became a salient concern to the Census Bureau following the unanticipated decrease of 10 percentage points in the short form mail response rate in 1990 from the 1980 census. The SQT was designed to determine the effects on response rates of asking fewer questions, using more "respondent friendly" forms, requesting social security numbers and using an implementation strategy that relies upon multiple contacts by mail. The study was conducted using the 1990 census short form as the control and four experimental versions of the 1990 census short form. The four experimental questionnaires included a user friendly "booklet", a shortened "micro" form, a shortened form including social security number (micro/SSN), and a postcard- like form referred to as the "roster" form. By comparing rates of question nonresponse across form types (in cases where content is comparable), this evaluation supplements the mail response rate analysis by providing a glimpse at the within-form levels of response.