Abstract Preliminary strength tests on laboratory sedimented soils have been run to ascertain possible failure criteria associated with hydrostatic anchors. It has been previously shown by Brown and Nacci 2 in model studies utilizing sands that the holding capacity of such anchors is a function of the diameter to skirt embedment ratio, the difference existing between the anchor cavity pressure and the surrounding hydrostatic pressure, and the amount of dead weight associated with the anchor system. The mechanistic behavior or the soil was not considered at that time. Information pertaining to the soils behavior has now been sought in connection with hydrostatic anchoring. Triaxial extension tests were performed to simulate actual failure conditions. These tests performed on laboratory sedimented kaolinite clay and on Providence silt indicate that the Mohr-coulomb shear parameters for extension and compression differ with the shear strength in extension being less at a relative effective stress than that of compression. It was also found that the ratio of the undrained shear strength to the effective consolidation pressure averaged about 26 per cent less for extension than compression. Failure occurred at lower strains and also exhibited higher A-factors when subjected to extension. These results indicate that an understanding of the shear strength properties of the underlying soil may be of significant assistance in predicting ultimate pullout forces. Introduction The influence of an applied stress system on the strength behavior of soils and sediments has been well known for years (Ladd, Varallyay6), for this reason on very important jobs involving soils, the use of laboratory tests that will closely simulate field loading is recommended. Generally this involves decisions on the use of drained or undrained triaxial compression tests, simple shear tests or plain strain tests to simulate the compressive load that is being applied to the soil. The use of earth anchors which results in a reduction or unloading force is not nearly as common as compressive loads in soils and where anchors are used in civil engineering works it is generally in denser sands and gravels because the compressibility of softer silts and clays restricts the amount of restraint that can be mobilized. The Use of shallow, hydrostatic anchors in generally soft cohesionless or cohesive sediments has therefore proven to be in an area where there is a dearth of analytical or reliable empirical approaches for the design of earth anchors. Some experimentation with hydrostatic anchors in sand (Brown, Nacci2) in the laboratory on scale models has shown a type of shear failure somewhat similar to that under a footing load. The analytic approach used for uplift loads is therefore similar to bearing capacity downward force approach with the principle exception being the evaluation of soil shear strength parameters. Analyt Ical Approach A method used to calculate the pullout capacity of shallow anchors is to estimate the earth mass which will move in concert with the anchor.