It is generally encountered as a many-branched shrub that grows up to 3 m (10 ft) high, with narrow leaves and the yellow inflorescences (flower spikes) appearing from late summer to early winter.

Cylindrical in shape, they are composed of a central woody spike, from which a large number of compact floral units arise perpendicularly to it.

Up to 150 follicles develop, each covered in short fine fur which is initially pale brown but fades to green-grey and partly wears away.

However, neither botanist considered this to be a new species at the time, instead holding it to be an unusual mountain form of the locally widespread B. marginata.

[8] Willis named the species after Victorian plantsman Bill Cane who had alerted authorities to the existence of an unusual banksia that was distinct from B. marginata some years previously.

[5] At the time, a plant collected from Mount Fulton near Port Davey in South West Tasmania was thought to be B. canei, but it was later reassessed as B. marginata.

[11] In his 1981 monograph of the genus Banksia, Alex George noted that despite a superficial resemblance to B. marginata, its bare old cones and stouter foliage indicated a closer relationship to B. integrifolia and B. saxicola,[5] although it lacks the latter species' whorled leaf arrangement.

[11] A fossil species, B. kingii from the late Pleistocene of Melaleuca Inlet in southwestern Tasmania, has robust foliage and infructescence resembling those of B. canei and B. saxicola, and appears to be a recently extinct relative.

[12] The leaf of a much older fossil species Banksieaephyllum acuminatum from Oligocene deposits in the Latrobe Valley closely resembles B. canei in shape, anatomy and vein pattern.

[13] The current taxonomic arrangement of the genus Banksia is based on botanist Alex George's 1999 monograph for the Flora of Australia book series.

He did place the two subalpine taxa (B. canei and B. saxicola) at the end of the sequence as he thought they were the most recently evolved species, since he considered the group to have a tropical origin and B. dentata to be the oldest lineage.

[20] Salkin described four populations (topodemes), each found in granite-based rocky soils in subalpine regions, and isolated from one another by wide river valleys.

[25] Animals recorded foraging among the flower spikes of B. canei include the yellow-tufted honeyeater (Lichenostomus melanops), bees, wasps and ants.

[7] Banksia canei lacks a lignotuber and appears to regenerate from bushfire by seed, although its response to fire has been little studied.

[8] Plectronidium australiense is a species of anamorphic fungus that was recovered from a dead branch of B. canei at Healesville Sanctuary and described in 1986.

[3] Although grown successfully in England and tolerant of temperatures to −12 °C (10 °F),[27] Banksia canei has a reputation of being difficult to keep alive in Australian gardens.

Salkin proposed this was necessary so that seed released in a summer or autumn bushfire would lie dormant over the winter months before germinating in the spring.

In January that year, two seedlings from the Wulgulmerang population displayed deeply lobed (pinnatisect) leaves and a prostrate habit.