Abstract: The evaluation of the efficiency of an earthquake prediction is based on statistical evaluation of the efficiency of middle-short term precursor prediction. Our study comes up with the evaluation method of "Error from target center" based on the concept of "error"("target practice" model), which takes the objective point of three factors to predict as center, the closed area circled by semi-precision prediction scale as target standard point, acceptance-error scale as target radius. If the distance between the real earthquake and the objective point(error) is less than the semi-precision scale, prediction hits the target center, i.e. realizing the semi- precision prediction(The score is 1). Then extending gradually to the acceptance-scale, exceeding the scale means "off-target"(score 0). According to present actual level of prediction, and considering the three types of prediction, the mid-term, short-term and short and impending prediction, and the difference of prediction magnitudes, we give the standard of semi-precision scale and acceptance-error scale. The cases have verified that method is efficient and available.

Key words: Earthquake prediction, Efficiency evaluation, Error from target center