Frequency estimation of sinusoidal signals is a fundamental question in systems theory that has various engineering applications for instance in active noise and vibration control in helicopters, disk drives, and magnetic bearings. Sinusoidal disturbances are narrow band and therefore their identification and cancellation in control systems is possible. In most cases the frequency of sinusoidal disturbances is unknown and since the sinusoidal disturbance cancellation methods are based on known frequency, frequency estimation in such cases is mandatory. In this work, iired from induction motor model, a frequency estimation method, namely, induction motor frequency estimator (IMFE), has been introduced. The proposed method uses transient differential equations of induction motor which show the relationship between rotor speed and total flux linkage of windings. This new frequency estimator has advantage of generation sinusoidal signal, its 90 degree phase shift and a sinusoidal signal with 45 degree phase shift beside. Local asymptotically stability analysis of IMFE is carried out based on the concept of integral manifold of slow adaptation and averaging theory. Simulation results show that the proposed IMFE has better performance compared with second order generalized integrator frequency locked loop (SOGI-FLL) which a well known frequency estimator. Then a sinusoidal disturbance cancelation method with unknown frequency based on the proposed frequency estimator has been developed. The proposed method consists of IMFE and an algebraic control law. The proposed algorithm has been compared with adaptive feedforward cancellation (AFC) algorithm. The AFC algorithm has a first order dynamic to produce control signal, and the proposed algorithm produces control signal without increasing the order of the dynamic of total system. Also, the proposed method need not direct use of estimated frequency in order to reproduce orthogonal sinusoidal signal because of generation the orthogonal signals at defined outputs of the IMFE. In the proposed method by estimating orthogonal sinusoidal signals, we directly cancel disturbance in control system. The plant under control is assumed that is stable, known, linear and time invariant. Simulation results show that the proposed cancellation method has faster parameters convergence speed and thus faster disturbance cancellation compared with AFC method. Finally, our proposed algorithm is modified to a disturbance cancellation for unknown plant using adaptive laws for estimation the real and imaginary parts of the plant frequency response at the frequency of the sinusoidal disturbance. Key Words: Adaptive feedforward cancelation, sinusoidal signal, orthogonal sinusoidal signal, frequency, averaging theory, induction motor model