The Dactylis glomerata L. species complex is a genetically diverse group of naturally occurring diploid (2n = 2x = 14), tetraploid (2n = 4x = 28), and hexaploid (2n = 6x = 42) subspecies. The diploid subspecies tend to be geographically isolated and genetically distinct. Each subspecies probably possesses alleles for adaptation to specific environmental regimes. The objective of this research was to evaluate tetraploid progeny from 2x–4x intersubspecific crosses for forage yield and quality relative to their tetraploid parent. Progeny were developed by two stages of hybridization; intra- and inter-subspecific hybrids were generated from 2x–2x crosses, after which selected diploid progeny were mated, as females, to adapted tetraploid genotypes. From 14 2x–4x crosses, utilizing diploids known to produce 2n eggs, 59 tetraploid progeny were identified. These tetraploid progeny and their parents were vegetatively cloned and established in pots in two greenhouses. Forage was harvested from two replicates of each plant in each greenhouse for a total of six cuttings. Vigor, in vitro dry matter digestibility (IVDMD), and neutral detergent fiber (NDF) were determined. Parent–progeny regressions indicated evidence of heritable variation for vigor and IVDMD, but not for NDF. Favorable heterosis (relative to the tetraploid parent) on a family mean basis was significant only for IVDMD in one family. When selection among progeny within families was possible (in six families), favorable transgressive segregants were observed for all traits in at least one family and in a total of five families. The favorable performance of some tetraploid progeny, combined with the diverse nature of the species, indicate that improvements to cultivated orchard grass are possible. Selection at both the diploid and tetraploid levels will be necessary to achieve these goals.Key words: Dactylis glomerata L., gametic nonreduction, heterosis, heterozygosity, orchard grass, polysomic polyploids.