An adaptive output feedback dead-zone compensation scheme is designed for systems with an unknown dead-zone at the input of an nth-order smooth nonlinear dynamics in the output-feedback canonical form. The proposed adaptive controller employs an adaptive dead-zone inverse to cancel the dead-zone and uses a back-stepping design for adaptive output feedback control. It has a new state observer parametrization that is needed to handle the dead-zone uncertainty and a new robust control law that is suitable for the parameter projection needed to implement an adaptive dead-zone inverse. The adaptive dead-zone compensation design ensures closed-loop signal boundedness and improves system tracking performance.