A moving picture encoding apparatus divides a moving picture signal into key frames and non-key frames, encodes the key frames, predicts the non-key frames from the encoded key frames, determines a final encoding rate for each non-key frame, encodes each non-key frame, and outputs an error-correcting code at the final encoding rate as encoded data of the non-key frame. To determine the final encoding rate, the number of error-correcting bits needed to correct prediction errors is estimated by a first method and one or more second methods. The second methods are used to decide whether to use the number of bits estimated by the first method as the final encoding rate, or whether to supplement the estimated number of bits by adding an encoding rate correction. As a result, the encoding rate is not supplemented unnecessarily and the encoding efficiency is improved.