Non-exclusion from paternity of alleged man can be quantified by pre-test and post-test parameters. Author advocates the view that expression of the power of forensic expert evidence by likelihood ratio, which compares probability of data under two hypotheses, is the best approach from both the point of forensic mathematics and judiciary practice. Advantage of likelihood ratio is described and demonstrated on microsatellite typing examples.

        Key words: paternity testing, likelihood ratio, Bayes theorem.