In sample survey design, information about the sampling units which is supplementary to the characteristics under investigation in the survey. Such information may be used for stratification, for the determination of the probabilities of selection, or in estimation. When used in estimation it provides the basis for estimators based on ratios or regression. For example, in a survey of business firms, supplementary information on, say, gross turnover provided by a previous census may be used either in the sample design or to improve the efficiency of sample estimates.