In this paper, a sample design is studied for 2000 forestry management survey of five types forestry , tree felling, gathering of pine mushroom, growing of nut trees, growing of wild flowers, and lumbering industry. We introduce population stratification and a modified stratified cut-off sampling which deal with determination of sample size, sample allocation, and estimation of total and variance of estimator. Substitution of sample units and imputation of nonresponse units are discussed for reducing the nonsampling errors.