A method of comparing and identifying two patterns is proposed based on the accidental coincidence probability of the patterns' features. The probability is calculated under the assumption that there is no correlation between the patterns. This approach makes it possible to overcome the difficult problem of estimating the probabilities of features missing and spurious features occurring. Applying this method to fingerprint identification, we show that it can be used to control the probability of imposter acceptance. Therefore, this method is suitable for security applications.