The need for timely and reliable agricultural information such as crop production estimates, livestock inventories and socio-economic data have become more important in decision-making process at international and national levels in almost all countries with the global shift towards market economies. Increasing population pressure may result in food scarcity especially in developing countries. Therefore, reliable crop area estimations are very important factors in food security. Moreover, cost is a very important consideration when resources available for statistical work are meager. Planning and conducting a sample survey requires a sound system in the organization especially when some agricultural data are concerned. As the staffing requirement is much less, it is possible to use well trained and motivated enumerators, specialized equipment and exercise, better and more intensive supervision, all of which are expected to improve the quality of the data collected and the accuracy of results. Based on this exercise it seems that every district should have the first round of surveys of this type to be used as the pilot to decide the number of clusters to be used and the cluster size for the particular district. Accordingly the decided sample design will help to provide estimates at a desired level of accuracy with minimum cost and time. Crop Yield Forecasting and Agro meteorology under the ASIRP Project of the Department of Agricultural Extension in Satkhira district used a large sample for estimating total area under Amon crop. Since there was enough field staff of DAE, collecting a large sample was not a problem for that project. But further work in this district might not have enough budgets to collect such a huge data. Moreover, there remains a risk of sampling error with large sample. So there is a need for further analysis on this data to examine the effect of various sizes of sample and sampling unit on the estimate of the parameter of interest. This analysis tries to do so. It may be helpful for further similar analysis in Satkhira district in determining the sample size with the size of sampling unit.