Technical Abstract: In many major perennial forage species, genomic tools and infrastructure development has advanced enough that their utilization in marker assisted selection (MAS) can be cheaply explored. This paper presents a paternity testing MAS in diploid red clover (Trifolium pratense L.). Utilizing individual plant phenotypes, known maternity, and molecular marker determined paternity, paternal and maternal breeding values are calculated and selection on both parents is accomplished. Paternity testing MAS is demonstrated in three red clover breeding populations utilizing permutation based truncation selection, for a biomass-persistence index trait. Permutation based truncation selection is accomplished by ranking parents based on 80% of total progeny per permutation. Parental rankings are then used to select among the remaining 20% of total progeny and average selection gains across all permutations are estimated. Paternity was determined from 11 SSR amplified in two PCR reactions. Paternity based selection gains alone were more than double selection gains based on maternity alone. Inexplicably, the estimated paternal halfsib family additive genetic variance was four to five times greater than the maternal halfsib family additive genetic variance. Paternity testing MAS is implementable in other diploid forage species and allopolyploid forage species with one diploid genome and corresponding genome specific molecular markers.