A network of laser interferometer gravitational waves detectors spread across the globe is currently running and steadily improving. After complex data analysis from the output signal of the present detectors, astrophysical results begin to emerge with upper limits on gravitational wave sources. So far, however no direct detection has been announced. To increase the sensitivity of current detectors, a second generation of interferometers is planned which will make gravitational wave astronomy a reality within one decade. The advanced generation of interferometers will represent a substantial upgrade from current detectors. Especially, very high optical power will circulate in the arm cavities in order to reduce by one order of magnitude the shot noise limited sensitivity in high frequency. However, the theoretical shot noise limit will only be achieved after implementation of complex thermal lensing compensation schemes. Thermal lensing is direct consequence of the residual optical absorption inside the substrate and coating of the test masses and could have tragic consequences for the functionality of the interferometer. The Australian Consortium for Interferometric Gravitational Astronomy (ACIGA) in collaboration with LIGO will run a series of high optical power tests to understand the characteristics and effects of thermal lensing. During these tests, techniques to compensate thermal lensing will be experimented. This thesis mainly focused on the first high optical power test in Gingin, Australia. The first test will consist of a Fabry Perot cavity with the sapphire substrate of the input mirror inside the cavity. Due to the high optical circulating power a strong convergent thermal lens will appear in the input mirror substrate. Because of the presence of the thermal lens inside the cavity, the size of the cavity waist will be reduced and the cavity circulating power will decrease. Simulations using higher order mode expansion and FFT propagation code were completed to estimate ways to compensate strong thermal lensing for the Gingin first test. The term `strong thermal lensing? is used because the thermal lens focal length is comparable to the design focal length of the optical components. The expected performance of a fused silica compensation plate is presented and advantages and limits of this method are discussed. Experimental results on small scale actuators which can potentially compensate thermal lensing are detailed. The knowledge gained from these experiments was valuable to design the real scale compensation plate which was used in the first Gingin test. This test was carried at the end of 2005. The thermal lens due to 1 kW of optical power circulating in the sapphire substrate was successfully compensated using a fused silica plate. Yet, thermal lensing compensation may only be required for room temperature advanced interferometer. Indeed, we showed that cooling the interferometer mirror to cryogenic temperature can eliminate the thermal lensing problem and also substantially decrease the mirror thermal noise.