A problem related to the estimation of population mean on the current occasion based on
the samples selected over two occasions is investigated. Some classes of estimators for estimating
the population mean at the current occasion in two-occasion successive (rotation)
sampling have been proposed. Properties of the proposed classes of estimators have been
studied. Optimum replacement policies are discussed. Estimators in the proposed classes
are compared with (i) the sample mean estimator when no information is used from the
previous occasion (ii) the optimum estimator which is a linear combination of the means of
the matched and unmatched portions of the sample at the current occasion and (iii) a chain
type regression to ratio estimator when auxiliary information is used at both the occasions.
Empirical comparisons are shown to justify the propositions of the estimators and suitable
recommendations are made.