Abstract: A linear constrained least squares constant modulus algorithm (LSCMA) is a convergent and steady blind beamforming algorithm used in array signal processing. When adaptive arrays are applied in practical situations, the performance of a linear constrained LSCMA is known to undergo substantial degradation in the presence of even slight mismatches between the actual and presumed array responses to the desired signal. To account for mismatches, a novel robust constrained LSCMA algorithm based on worst-base performance optimization was proposed. The proposed algorithm estimates direction of arrival (DOA) of the actual signal from the observations by using a posterior probability density function, and analyzes theoretically the array output performance. Robust constrained LSCMA provides excellent robustness against uncertainty in the DOA of the actual signal and makes the mean output array SINR consistently close to the optimal one. Computer simulation results are presented to support better performance of the proposed algorithm compared with the linear constrained LSCMA algorithm.