Because the disturbances which govern the dynamical response of a structure cannot be precisely measured, and the system itself has many uncertainties, the development of control strategies that are implementable and that can accommodate uncertainties and imprecision are becoming a critical and challenging work. PID adaptive controller based on RBF Neural Networks Identifier is developed for structural control in this paper. The combined controller includes PID neural network controller and an identifier based on RBF neural networks. It was implemented on linear single degree of freedom system representation of structures subjected to external disturbances based on the El Centro (1940), Hachinohe (1988), Kobe (1995) and Northridge (1994) earthquake loadings. It is demonstrated that the neuro PID adaptive control method can effectively suppress the vibration of structures.