In several food processing and chemical industries, liquid is pumped and kept in interrelating coupled tanks. However, automatic regulation of the liquid level and flow control between these tanks is a challenging problem because of the complexity and high non linearity of such system. This paper deals with the liquid level control of two horizontal coupled tanks system. A comprehensive comparative study is made for most popular sliding mode control (SMC) algorithms found in literature, namely Proportional-Derivative Sliding Mode Control (PD-SMC), Proportional-Integral-Derivative SMC (PID-SMC), Fractional Order SMC and finally dynamic SMC. Special emphasis is put on the effect of the sensor noise on the controller performance. Simulated experiments including robustness to variation in plant parameters and step input disturbances are made. Control algorithms parameters are selected to optimize designed performance indices by using MATLAB optimization toolbox. Simulation results reveal that dynamic SMC is superior to other control algorithms in the presence of sensor noise and has a significant reduction in the actuator chattering phenomenon.