A novel continuous auto-calibration technique was presented for redundant strapdown inertial navigation systems. Principles of the proposed technique were based on strapdown system error models, which indicated that velocity errors from the system could contain information related to inertial sensor errors when rotated along a predefined trajectory. A novel calibration trajectory was designed. All deterministic errors could be accurately estimated through Kalman filtering techniques. Simulation platform for redundant strapdown system was designed to validate efficiency of this technique. Results show that the novel calibration technique outperforms conventional multi-position technique considering its simplicity and precision, which is of great importance for practical applications.