Abstarct: Physical and mechanical properties of construction materials is a fundamental issue that affects all phases of a civil or mining project such as designing, building, stability analysis, maintenance and failure prediction. Exact information of physical and mechanical properties of rock is a vital practice. In situ physical and mechanical properties defining has difficulties such as costly and long experiments and difficult coring. Consequently, evaluating rock properties with non-destructive methods have been increasingly demanded. One of the most important non-destructive methods is UPV which is used for rocks in experimental and field conditions. Many researchers have been used this method as a fast and economical tool for estimating elastic and mechanical properties of rock and presenting empirical relations between UPV and physicomechanical properties of different rock types. Stress state is another important factor that should have been specified during different phases of civil or mining projects. Experimental tests on rock samples approved UPV method efficiency for determining stress state. Ultrasonic pulse velocity is affected by the current and past stress which was tolerated by the rock. So defining a logical relation between stress state and ultrasonic pulse velocity makes it possible to obtain useful information about conditions of structural stability, failure prediction and destruction prevention. A relationship between ultrasonic pulse velocity and stress state is presented in this study. To achieve this aim, the ultrasonic pulse velocity test in different stress levels was defined on gypsum and cement cubic samples. Optimal mixing ratio achieved by uniaxial compressive strength test on standard cylindrical samples (diameter Nx) with different mixing ratios. So, 22 cubic samples were made with dimensions of 10 x 10 x 10 cm3 and 0.5 of water to cement ratio. To determine the variation of axial and lateral strain, surfaces were smooth with adhesive polishing. Two 60 mm strain gauges were glued on the sample with the direction of parallel and perpendicular to the axis of loading. The specimen was loaded under release and different stress conditions with predetermined fixed intervals to failure. In each stress level, wave transit time were measured using 55 kHz vertical probes in continuous mode and perpendicular to the loading axis, lateral and axial strain was recorded using strain gauges installed orthogonal on the sample. Finally, the logical relationship between these two characteristic information are presented baxsed on the stress-strain curve of samples as well as the ultrasonic pulse velocity in them.