Adaptive Cluster Sampling is a sampling approach in wich the selection of sampling units in any stage of sampling depends on the information the units already selected in the previous to stages. The usual unbiased estimators in adaptive cluster sampling are simple to compute but do not necessarily use all information thet has been gathered. In this thesis. Adaptive cluster sampling in wich initial sample of units is selected by simple random sampling is introduced. Then usual and Rao-blackwell estimators and a sample of application of this design and Rao-Blackwell estimators are described.