The output feedback algorithm for dynamic plants with compensation of parametric uncertainty, external disturbances and measurement noises is synthesized. The plants are described by a nonlinear system of differential equations with vector input and output signals. Unlike most existing control schemes in this paper the dimensions of the measurement interference and the output signal are equal, the sources of the signals of disturbances and disturbances are different, parametric and external disturbances can be present in any equation of the plant model. For simultaneous compensation of disturbances and measurement noises it is proposed to consider two channels. On the first channel a part of the measurement noises will be estimated which will allow partial recovery the information about the plant noisy output. On the second channel the disturbances will be compensated. Thus, at least two independent measurement channels are required for simultaneous compensation of disturbances and measurement noises. Sufficient conditions for calculating the parameters of the algorithm in the form of solvability of the linear matrix inequality are obtained. It is shown that the equation of a closed-loop system obtained on the basis of the proposed algorithm depends on the disturbances and the smallest component of the measurement noise. However, if the smallest component cannot be identified a priory, the results of the transients depend on the component of the noise that will be selected in the synthesis of the control system. Thus, unlike most existing control schemes, where the equation of a closed-loop system depends on disturbance and noise, the resulting algorithm provides better transients, because they do not depend on the entire noise vector, but only on its smallest (one) component. The simulations for a third-order nonlinear plant and the synchronization of an electrical generator connected to the power grid are presented. Numerical examples illustrate the effectiveness of the proposed scheme and the robustness with respect to random components in the noises and disturbances.