The invention relates to a star sensor and gyroscope data fusion method suitable for an attitude maneuver working condition. The star sensor and gyroscope data fusion method comprises the following steps: determining a star sensor measurement gain coefficient based on the principle of least squares; obtaining a current beat attitude through one-step recursion based on a last beat estimated attitude and a current gyroscope measurement angular velocity; fusing star sensor measurement attitude data and gyroscope recursion attitude data. According to the star sensor and gyroscope data fusion method disclosed by the invention, star sensor data and gyroscope integral data are fused by adopting a least square method, attitude estimation accuracy can be increased by utilizing the characteristic of small gyroscope integral attitude noise, a non-divergent attitude is kept by utilizing the characteristic of accurate star sensor measurement values, and the problem of inaccurate attitude estimation during an attitude maneuver period of an existing EKF (Extended Kalman Filtering) algorithm is solved.