The inherent inseparability of a cost-effective Soil Sampling Quality Assurance/Quality Control (QA/QC) Plan from the objectives of a soil monitoring program is emphasized. Required precisions and confidence levels for the data cannot be defined until the decisions which will be made on the basis of the data are clearly stated and the consequences of making Type I (false positive) or Type II (false negative) errors are weighed. Statistical considerations are presented with special attention to analyses of variance of soil monitoring data, methods of calculating required numbers of soil samples to achieve desired precisions and confidence levels, possible applications of Kriging, and assignment of control limits to QA/QC data. The value of an exploratory or preliminary study to the cost-effective achievement of both the soil monitoring objectives and the objectives of the Soil Sampling QA/QC Plan is strongly emphasized. The value of developing a hypothetical model to estimate the distribution in space and time of soil pollutants and thus to assist in the design of the monitoring network is discussed. Methods for determination of the number and locations of soil sampling sites; sample collection methods and procedures to include frequency of sampling; sample handling to include labeling, preservation, preparation for analysis, and transport; together with QA/QC aspects are presented and discussed. Finally, the importance of systems audits and training to the achievement of soil sampling QA/QC objectives is presented and discussed.

"EPA-600/4-84-043." "PB84-198621." "May 1984." "Cooperative agreement CR 810550-01, Kenneth W. Brown, Project Officer." Includes bibliographical references (pages 89-92).