Engineers of the Natural Resources Conservation Service (formerly the Soil Conservation Service), USDA, instituted a testing program to determine the reliability and reproducibility of several types of field density test methods. Tests were performed at several embankment construction sites in the Southeastern United States. Several different soil types were encountered in the embankment projects. On test pads constructed under normal field conditions, researchers measured the compacted dry density and water contents by the nuclear gage, sand-cone, drive-cylinder, and clod tests. The research compares the results of the various tests and discusses the conclusions of the researchers on reliability and accuracy of the various tests. Conclusions of the studies include important aspects of each lest that were found to explain variations and errors in the tests that were performed. The paper emphasizes errors common to the tests evaluated. Recommendations based on the studies of the NRCS are included. Advantages and disadvantages of methods are discussed. Data from the study sites is included to illustrate repeatability and comparisons of the various methods to each other. Data from the study sites are included to illustrate repeatability and comparisons of the various methods with each other.