For aircraft navigation in unknown environments, a velocity and height estimation method based on vision/inertial integrated navigation is proposed. An extended inertial navigation state equation was formulated, which contains several aircraft’s positions at latest several imaging times, and a linear vision measurement equation based on two view epipolar geometry constraints was adapted to correct inertial velocity error using a Kalman filter. And then, the coordinates of ground features were estimated by the stereo vision method, which was used to estimate the height of the aircraft. Simulation has been implemented by using the typical cruise trajectory of the aircraft, and the result shows that our method works well for correcting the velocity and height errors, which enables the velocity and height not to drift with time. As a result, it can also suppress the position error of the aircraft.