Herb, stem erect, simple or branched, sometimes not higher than 40 cm, or plant scandent or climbing, up to several meters long. Leaves alternate, sometimes opposite or verticillate and clustered, sometimes clustered above the middle of the stem, sessile, base sometimes sheathing the stem, ranging from linear, elliptic-lanceolate, elliptic to ovate, 6-17.5 cm long, 0.4-5 cm wide, apex acute, acuminate, falcate or ending in a tendril. Flowers solitary, axillary, sometimes terminal, different shades of yellow, orange, red, crimson, purple/mauve stripes or fading purple, often bicolored; pedicel erect, recurved apically, 3.5-18 cm long. Perianth sometimes at the base connate into a short tube, up to 2 mm long; perianth segments (strongly) reflexed, base clawed, narrowly elliptic-linear, oblong-lanceolate, sometimes ovate to obovate, 29-85 mm long, 4-25 ( -38) mm wide, the margins sometimes crisped, apex acuminate to acute, sometimes falcate. Filaments filiform, sometimes flattened, 10-45 mm long; anthers straight to curved, 5.5-15 mm long. Ovary 4-13 mm long, 1-5 mm wide; style 9-50 mm long. Capsule 37-50 mm long, 10-14 mm in diameter; seeds 4 mm.
Gloriosa undoubtedly exhibits a great variety of forms in the habit, leaf shape and shape and structure of the flower. The differences in habit and floral form have been used to distinguish several "species" in the past based upon specimens collected in Africa. The absence of the leaf tendril and the variation in leaf shape have also been used in the differentiation of species. These characteristics have proven not to be consistent even in individual plants. For example, the presence or absence of leaf tendrils in the erect forms depends entirely on the growth of the plant, and it is frequently only the uppermost leaves of some tall plants which develop tendrils. Plants that develop them one season may do not so in the next. It is not known whether it is possible for an erect, non-climbing plant to go into a climbing habit. Therefore I have decided that the characters used earlier to separate a number of "species" are not adequate for separating taxa, and all material from East Africa is included in G. superba L.
The dwarf growth habit of specimens from Northern Kenya, Ethiopia and Somalia is still recognized as being worthy of taxonomic recognition, but only as a variety. This growth habit seems to be restricted to arid regions.