Due to the uncertainty of the maneuvering platform attitude, the current traditional height estimation algorithms based on signal echo or relying on the platform attitude will be limited, which leads to accuracy degradation. To address this problem, this paper proposes a beam-scanning-based height estimation method for phased-array altimeters, aiming at realizing high-accuracy altimetry and attitude angle inversion with unknown platform attitude. The method establishes a geometric relationship model among height, attitude angles, slant ranges and beam directions, uses the phased-array antenna to quickly scan multiple beam directions to obtain the echo and estimate the corresponding slant range. Based on the geometric model, a system of equations is constructed, which is solved by the least-squares method to invert height and attitude angles. This method overcomes the limitation of the traditional algorithm by using the phased-array to scan multiple directions, and effectively reduces the influence of the slant range error on the accuracy through the least squares optimization. Simulation experiments show that the relative errors of height estimation of the proposed method are all less than 1%, and the pitch and roll angle inversion errors are all less than 1° in a variety of platform attitude scenarios, which indicates that the proposed method has high accuracy and stability. Compared with the traditional altimetry algorithms, the method proposed in this paper not only can measure height with high accuracy in a stable way, but also has a higher computational efficiency. In addition, the measured data further verify the effectiveness of the algorithm, which provides a reliable solution for the height estimation of maneuvering platforms.