When adaptive arrays are applied to practical problems, the performance degradation of adaptive beamforming techniques may become even more pronounced than in the ideal case because some of underlying assumptions on the environment, sources, or sensor array can be violated and this may cause a mismatch between the presumed and actual signal steering vectors. In the practical environment, complete knowledge of signal characteristics is not available and the environment is time varying. In these cases, the recursive algorithms to robust adaptive beamforming are required. In this paper, we propose robust constrained-LMS algorithm based on constrained-LMS algorithm and explicit modeling of uncertainties in the desired signal array response, which belongs to the class of diagonal loading approaches. Our proposed robust constrained-LMS algorithm provides excellent robustness against the signal steering vector mismatches, enhances the array system performance under nonideal conditions and makes the mean output array SINR consistently close to the optimal one. Computer simulations demonstrate a visible performance gain of the proposed robust constrained-LMS algorithm