Well integrity is one of the most important concepts in the upstream oil and gas industry. No matter whether a well is in drilling phase or has been already completed, the well integrity should be maintained. It is a multidisciplinary approach and has always been a serious challenge for major oil companies, regardless of operator or contractor. Therefore, all various departments in an oil company including drilling, completion, production, and plug and abandonment must cooperate and apply different methods, software and hardware, in order to maintain the well integrity for each operation of each section of each well. This research work aims to perform ultrasonic measurement as a non-destructive evaluation method (NDE) to estimate the compressive strength of oil well cement with different densities and recipes at several ages. Therefore, the main project’s target is to find a correlation between these two mechanical properties of oil well cement, the uniaxial compressive strength and the ultrasonic wave velocity, by inducing ultrasonic waves into the specific cement with unique composition at the known age. Using the correlation allows estimating the compressive strength of the cement in various densities and formulations other than the ones used in this project. As to the methodology, six cement compositions with different recipes have been chosen for the linear measurement tests in which the ultrasonic wave velocity is correlated to the compressive strength of oil well cement. Additives used to prepare a wide density range from 11.0 to 16.0 ppg include barite, bentonite, and 3M glass bubble. Different regression methods have been tested only best, with the highest R2 of 0.96 chosen to estimate the strength for 2D models. The factors considered in the model as input include water-to-cement ratio, density, ultrasonic wave velocity, and age. The result shows that the composition with higher density, made by barite, is more resistant against being fractured and can convey an ultrasonic wave faster, while the lightweight cement, bentonite, and glass bubble made, has a lower compressive strength and a slower ultrasonic wave velocity. Moreover, a function model is created based on the correlations between compressive strength, ultrasonic wave velocity, age periods, and density. The model's inputs are ultrasonic wave velocity, age, and cement density, while the compressive strength is the model’s output. Moreover, by using the acquired data, different correlations among (UCS, density, age) and (UCS, UWV) and (UCS, age) and (UWV, age) have been shown all of which have accuracy above 0.96 R2. Among the 2D correlations, UWV showed the best power fit for all cement samples. Eventually, by employing the 2D correlations and their generated coefficients, a rigorous analytical approach was given to define a generic 3D model of UCS as a transcendental function of several variables, including the density, age, and UWV. The function obtains the accuracy of R2=0.77. The modeling was a rigorous analytical approach. In case more data is generated in future work, the accuracy of the model can be improved. Another approach to improve the inherent accuracy could be employing machine learning techniques to get the best fit in transcendental functions rather than algebraic functions.