This article investigates the fault-tolerant control problem for stochastic switched interval type-2 (IT2) fuzzy time-delayed uncertain systems based on unknown input observer synthesis, which can avoid uneasy measurement on the time derivative of output, and estimate unavailable or partially measurable states, including sensor and actuator faults accurately. First, a desired fuzzy observer is designed to ensure the observer-based dynamic error system mean-square exponentially stable with sufficient condition of a strict (l,ħ,℘) - [Formula: see text]-dissipative performance, which is a unified framework of passivity, and H∞ provides results with less conservativeness. Then, we concentrate on stability analyses on dissipativity-based switched IT2 fuzzy systems with stochastic perturbation through linear matrix inequalities, Lyapunov function, free-weighting matrices, and average dwell time, discussing it according to different values of disturbance. Finally, simulation examples are listed to account for availability and effectiveness of the research methodology.