The adaptive interval type-2 (IT2) fuzzy output feedback control problem is studied for a single-phase photovoltaic grid-connected power system. The equivalent resistors of the inductors in the system are unknown and the part states are not available. Interval type-2 fuzzy logic systems (IT2FLSs) are utilized to approximate the uncertain nonlinear dynamics, and an IT2 fuzzy state observer is designed to estimate the unavailable states. By introducing a command filter method and using a backstepping control design technique, an IT2 fuzzy output feedback control scheme is investigated, in which the constraint conditions of pulse width modulation are ensured via mean-value theorem. It is proved that all the variables of the closed-loop photovoltaic system are uniformly ultimately bounded. The simulation and comparison results demonstrate the validity of the proposed control scheme.