This is a report on the first effort to formulate an effective question for determining the ethnicity of respondents to Great Britain's 1981 census. Procedures used to design and administer the trial questionnaire are outlined. Successes and problems in contacting and gaining the cooperation of various ethnic groups represented in the study sample are analyzed. The quality of answers given and respondents' difficulties with the questionnaire are described. Also discussed are respondents' personal views about the desirability of asking about race or ethnicity in censuses and social science surveys. The results from this trial survey are compared with the current method utilized, which infers ethnicity from parental birthplace data. Appended to this report are a description of the methods by which the sample was selected, a fieldwork report, instructions for interviewers, and forms used in the survey. (GC)