Plants annual or perennial; usually synoecious, sometimes monoecious or dioecious; habit varied. Culms usually annual, sometimes becoming somewhat woody, internodes solid or hollow. Leaves sometimes conspicuously distichous; sheaths usually open; auricles absent; abaxial ligules usually absent, sometimes present as a line of hairs; adaxial ligules membranous, often ciliate with cilia longer than the membranous base, sometimes not ciliate; blades not pseudopetiolate; mesophyll usually radiate; adaxial palisade layer not present; fusoid cells absent; arm cells absent; Kranz anatomy present; midrib simple; adaxial bulliform cellspresent; stomatal subsidary cells dome-shaped or triangular; bicellular microhairs present, usually with a short, wide apical cell; papillae sometimes present. Inflorescences ebracteate, paniculate, racemose, or spicate(occasionally a single spikelet), if paniculate, often with spikelike branches; disarticulation usually beneath the florets, sometimes at the base of the panicle branches. Spikelets usually bisexual, usually laterally compressed, with 1-60 florets, distal florets often reduced. Glumes usually 2, shorter or longer than the lemmas, sometimes exceeding the distal florets, lower or both glumes occasionally missing; lemmas lacking uncinate hairs, sometimes awned, awns single or, if multiple, the bases not fused into a single column; anthers 1-3; ovaries glabrous; styles 2, separate throughout, bases close. Caryopses often with a free or loose pericarp; hila short; endosperm hard, without lipid; starch grains simple or compound; haustorial synergids absent; embryos usually large relative to the endosperm, not waisted; epiblasts usually present; scutellar cleft present; mesocotyl internode elongate; embryonic leaf margins usually meeting, rarely overlapping. x = (7, 8,) 9, 10 (12).

The subfamily Chloridoideae is most abundant in dry, tropical and subtropical regions. In the Flora region, it reaches its greatest diversity in the southwestern United States (Barkworth and Capels 2000). Almost all its members, and all those in the Flora region, have C₄ photosynthesis. Most employ the NAD-ME or PCK pathways, but Pappophorum utilizes the NADP-ME pathway.

The subfamily has been recognized, with essentially the same limits as here, for some time, although reservations have been expressed concerning its monophyly (Campbell 1985; Jacobs 1987; Kellogg and Campbell 1987). More recent studies, both morphological (Van den Borre and Watson 1997, 2000) and molecular (Soreng and Davis 1998; Hilu et al. 1999; Hsaio et al. 1999; Grass Phylogeny Working Group2001; Hilu and Alice 2001) support its recognition as a monophyletic unit. There is less agreement concerning the subfamily's closest relative, some studies pointing to the Arundinoideae (Grass Phylogeny Working Group 2001) and some to the Danthonioideae (Barker et al. 1995; Hilu and Esen 1993; Hilu and Alice 2001).

There is considerable disagreement concerning the tribal treatment within the Chloridoideae, the number of tribes recognized varying from two (Prat 1936) to eight (Gould and Shaw 1983). Hilu and Wright (1982, p. 28) concluded, on the basis of their morphological study, that the boundaries between most of the tribes in this subfamily are not pronounced. They noted that Savile (1979) reached the same conclusion from considering the host specificity of various pathogenic fungi.

More recent work supports Hilu and Wright's (1982) conclusion. Van den Borre and Watson (1997, 2001) recognized eight informal groups within the subfamily. Five of the groups were large, the smallest including around 133 species and the largest around 380. The other three groups, which correspond to the Orcuttieae, Pappophoreae, and subtribe Triodiinae, include 9, 42, and 54 species, respectively. The difference in size is of no concern; the fact that all three of the small groups are embedded within one of the five large groups, the Pappophoreae and Triodiinae in a group than includes Eragrostis subg.Eragrostis and the Orcuttieae in the group that includes Muhlenbergia, is disturbing. Van den Borre and Watson noted that part of the problem was that Eragrostis, and probably some of the other large genera, are not monophyletic.

Hilu and Alice (2001) recognized four clades within the Chloridoideae. Like Van den Borre and Watson, they found the Orcuttieae and Triodiinae to be monophyletic, although their placement within the subfamily was not clear. Unlike Van den Borre and Watson, Hilu and Alice found Pappophorum, and hence the Pappophoreae, to be polyphyletic.

The treatment presented here is conservative in recognizing the Orcuttieae and Pappophoreae as distinct tribes. It departs from most other treatments in merging all other North American taxa into a single tribe, the Cynodonteae. Consensus on how the Cynodonteae sensu lato should be broken up is unlikely to be reached until the generic limits of its members have been more thoroughly examined.