Many attempts to find a non-invasive procedure to measure the blood pressure locally in the body have been made. This dissertation focuses on the approaches which utilize highly compressible ultrasound contrast agents as ambient pressure sensors. The literature within the topic has been reviewed. From this, the appropriate pressure dependent acoustic properties of the microbubbles can be summarized to be the resonance frequency, the disappearance time, and the subharmonic response. During this thesis, the ambient pressure sensitivity of the subharmonic response has been investigated through simulations and initial experimental measurements. By simulations, a parameter study has investigated what mechanisms of the driving pulse are important to optimize the ambient pressure sensitivity when utilizing the subharmonic component. Investigating two different types of microbubbles clearly showed that two factors are important when striving for an optimum sensitivity. First, the amount of subharmonic energy reduction, when increasing the ambient pressure, is very sensitive to the acoustic excitation pressure. Second, the study also indicated that the amount of reduction in subharmonic energy is increased as the length of the excitation pulse is extended. To carry out measurements in the laboratory, an experimental setup has been established. As the focus has been on preparations for future in vivo measurements, the setup was designed to match a clinical situation. Under the current measurement conditions, this setup showed that the subharmonic component by itself cannot be used as an ambient sensitivity measure. Instead, a new technique looking at the ratio of the subharmonic energy to the energy of the fundamental component was used. Doing so, an ambient pressure dependent behavior of the microbubbles was observed, indicating this to be a more robust measure. When increasing the ambient pressure, the relation decreases linearly. Likewise, decreasing the ambient pressure makes the relation increase linearly. Although the approach seems to reduce factors like time dependency, a high standard deviation was still observed. This could be caused by several reasons and more measurements are needed to investigate it further.