The quadruple tank system is a challenging multi-variable model consisting of four interconnected tanks, two level sensors and two input pumps. This paper deals with the control and estimation of such a system. All the four levels of the system are controlled by H∞ filter based discrete sliding mode control (SMC). A nonlinear stable sliding surface is constructed using full state information, to enhance the overall performance of the closed loop plant. In this multi-input multi-output quadruple tank system, only the first two states are available for feedback and hence it is not suitable to use any full state feedback control methods directly. To circumvent this problem, the remaining two states, which are required for the SMC design are estimated using a robust H∞ filter. The efficacy of the proposed method is demonstrated by simulations.