Most soils that concern geotechnical engineering are in the state of partial water saturation. Current practice tries to predict engineering properties of cohesionless soils using data from tests on saturated specimens, regardless of the saturation in the field. Due to complexity of test setups and high technical requirements, unsaturated soil tests are not among the common equipment of soil mechanics laboratory. One of these problems is the existence of suction, which is a function of water content and affects the strength behavior of unsaturated soils. Procedures to keep the water content of the partially saturated specimens constant and homogeneous in conventional soil tests are not well-established. The exception to this is unsaturated test setups, which are costly, complicated and found only in research institutions, hence prohibiting the industry from keeping up with the developments in this field. This study explores simple modifications to conventional methodologies of triaxial and direct shear tests, with the ultimate aim of preventing temporal and spatial variability of specimen water content throughout test duration. For different modifications, specimens of each test are dissected at the end of the test, and water content profiles of the specimens are obtained.