Marine oil spills can cause serious environmental damages to natural resources and to those whose sustenance depends upon these resources. Unfortunately experience shows that even the best efforts have not prevented occasional occurrences of major accidents on the sea. As long as massive oil spills are probable, special techniques and equipments will remain essential to facilitate spill cleanup in coastal regions. Oil spill containment booms are the most commonly adopted techniques to collect and contain oil on the sea surface, or to protect specific areas against slick spreading. Recently, an anti-pollution boom called the Cavalli system, has been designed with the intention of preventing the spread of spilled oil by trapping it inside a flexible floating reservoir and improving the pumping operation by decreasing the reservoir surface, and consequently increasing the oil layer thickness. Although flexible barriers have become increasingly common as a cleanup facility, there is no more than inadequate elaborate knowledge about their behavior. According to an extensive literature review, most of existing researches, either physical or numerical, have been done for rigid barriers. The main motivation for introducing the present research project is to study the efficiency and operational limits of the Cavalli system. However, the objectives are not constrained to this particular case. The present investigation focuses on the behavior of flexible barriers containing spilled oil. Previous researches of containment booms, even for the case of rigid barriers, have been mainly carried out in calm water. Accordingly, the main concentration is devoted to the response of a flexible barrier in presence of sea waves. Both experimental and numerical approaches were pursued to evaluate the efficiency limits and behavior of flexible barriers. Two-dimensional experiments have been carried out in a laboratory flume 6.5 m long, 1.2 m deep, and 12 cm wide. Flexible and rigid barriers containing rapeseed oil were examined, with and without waves. As the first step, the behavior of a flexible barrier in currents without waves was studied and compared to that of a rigid barrier. The key challenge was to contain the oil behind a flexible barrier that can freely deform in the water flow. This could be achieved using a slitted side skirt on the boom where it faces the lateral wall of the flume. The failure mode observed for rapeseed oil was entrainment failure. The initial failure velocity of different experimental conditions was studied and an empirical relationship was suggested in order to assess the maximum permissible oil-water relative velocity as a function of barrier draft and oil characteristics. The geometrical characteristics of the contained slick were examined and empirical equations were proposed to predict the slick length and headwave thickness as a function of contained oil volume. The second and more significant step was to conduct experiments with a flexible floating barrier in presence of five different waves. The analysis focused on the relationship between the failure velocity and the wave parameters with an emphasis on the behavior of flexible barriers. Likewise, empirical equations were proposed for the prediction of the initial failure velocity and geometrical characteristics of the slick. A type of drainage failure, namely, surging drainage was observed in the presence of waves. It was shown that the wave steepness and oil layer thickness are the dominant parameters in such failure. It was noticed that by decreasing the wave period or increasing the wave height, interfacial waves became more aggressive and consequently failure initiated at a lower velocity. Flexible barriers were more sensitive to the variations of wave characteristics. Applying appropriate time and length scales, a critical wave period of 6 s and wave height of 0.5 m were proposed for the prototype. Accurate measurements of velocity profiles and flow patterns in the vicinity of barriers with different conditions by means of Ultrasonic Velocimetry Profiling (UVP) and Large-Scale Particle Image Velocimetry (LSPIV) methods provided a reasonable understanding of the hydrodynamics in the vicinity of the barrier. The characteristics of the headwave at the upstream end of the oil slick were deliberately compared to those of a gravity current. It was concluded that despite geometrical similarities, these two phenomena are quite diverse. Furthermore, the oil-water interface was traced by detecting the maximum ultrasonic echo intensity, and velocity profiles in water and oil phases were independently obtained. To enhance the understanding of the mechanisms associated with oil containment failure, numerical simulations of multiphase flow were carried out using FLUENT code, applying the finite volume method (FVM). Comparisons between the obtained flow pattern and velocity field derived from numerical simulations and precise experimental measurements confirmed the capability of the numerical model to simulate the multiphase flow. The turbulence wake downstream of rigid and flexible barriers was simulated with and without the presence of oil phase. The simulations revealed the effect of contained oil on flow pattern and consequently the drag force acting on the barrier. Simulations of a full-scale barrier proposed a drag coefficient, Cd, of 1.90 for rigid barriers. Contrarily a constant value for the drag coefficient cannot be attributed to flexible barriers, since its deformations do not allow it to form similar shapes at different velocities. Last but not least, comparing the drag force on a rigid barrier with that of a flexible barrier towed by the same velocity demonstrated the fact that the forces acting on the skirt could be appreciably reduced by allowing flexibility.