Brazilian presalt reservoirs comprise carbonate rocks saturated with light oil with different amounts of and excellent productivity. The occurrence of giant-size accumulations with such productivity generates the interest in production monitoring tools, such as time-lapse seismic. However, time-lapse seismic may present several challenges, such as imaging difficulties, repeatability, and detectability of small variations of reservoir properties. In addition, when assessing time-lapse seismic feasibility, the validity of Gassmann’s modeling for complex, heterogeneous carbonate rocks is arguable. Other questions include the pressure variation effects on the seismic properties of competent rocks. The effective stress is a linear combination of confining stress and pore pressure that governs the behavior of physical properties of rocks. Many applications assume that the effective stress for elastic-wave velocity is given by the difference between confining stress and pore pressure, whereas another common approach uses the Biot-Willis coefficient as a weight applied to the pore pressure to estimate the effective stress. Through a series of experiments involving ultrasonic pulse transmission on saturated core plugs in the laboratory, we verified the applicability of Gassmann’s fluid substitution and estimated the empirical effective stress coefficients related to the P- and S-wave velocities for rock samples from two offshore carbonate reservoirs from the presalt section, Santos Basin. We observed that Gassmann’s equation predicts quite well the effects of fluid replacement, and we found that the effective stress coefficient is less than one and not equal to the Biot-Willis coefficient. Moreover, there is a good agreement between the static and dynamic Biot-Willis coefficient, which is a suggestion that the presalt rocks behave as a poroelastic media. These observations suggest that more accurate time-lapse studies require the estimation of the effective stress coefficient for the particular reservoir of interest.