A total of 252 black and white trios (mother, child, and putative father) in which the alleged father was known not to be the biological father (BF) were generated from paternity studies. The likelihood of paternity (W) and the paternity index (P) were calculated on the basis of ABO, Rh, MNS, and HLA phenotyping. The results were then reviewed to evaluate the arbitrary limits proposed by a number of states in recent paternity legislation. Two hundred forty-four of 252 non-fathers (NFs) (96.8%) were excluded. Of the 152 white NFs, 147 (96.7%) were excluded; 97 of the 100 blacks (97%) were excluded. Eight NFs could not be excluded by the routinely performed tests. The W value for those eight trios ranged from 1.2 to 98.8 percent. Based on limits proposed in legislation, had the NFs actually been alleged fathers, as many as four of the eight of these men could have had the burden of proving that they were not the BFs.