A comparison of the HIS sample's characteristics with government census and survey data suggests that the HIS sample adequately represents the intended population. The levels of response error variance in the HIS measures compare favorably with those obtained in other personal interview surveys and are low enough not to seriously distort estimated relationships nor attenuate gains from the multiphase sample design. Respondent rule and level of question detail have a small effect on mean reported income and no effect on error variance. Error variance of income responses is lowest for the most recent year and increases for prior year responses. Mean and error variances for individual characteristics such as educational attainment are not affected if obtained from a proxy respondent rather than from the person himself. Reinterviews conducted by telephone within a short time of the original interview are deemed optimal. 134 pp. Ref.