Robyns (1963) proposed Pseudobombax grandiflorum var. majus based on specimens with longer pedicels and staminal tubes as well as wider calyces and petals than are found in the nominate variety. Analysis of the protologue and type material of this variety and the nominate one, along with a comprehensive examination of herbarium specimens, reveals that P. grandiflorum var. majus is also morphologically distinct from the nominate variety in having mostly 9–11 (vs. mostly 5) leaflets (7 leaflets rarely occur in both taxa), often 3–7-flowered cymes (vs. flowers solitary, rarely 2–3-flowered cymes), and 5-angulate (vs. 5-costate) fruits in cross-section. Moreover, P. grandiflorum var. majus has calyces that are urceolate (vs. cupuliform) and often lobed (vs. truncate).

In addition to these morphological differences, the two taxa can be distinguished by their distribution in Eastern Brazil. Pseudobombax grandiflorum var. majus inhabits semi-deciduous forests and occasionally granitic outcrops in wet forests whereas the nominate variety inhabits mainly coastal restinga vegetation. Thus, morphological evidence along with the parapatric distribution are sufficient to recognize Pseudobombax grandiflorum var. majus at species rank.

Pseudobombax majus can be distinguished from its sister species P. longiflorum (Mart.) A. Robyns (Carvalho-Sobrinho et al. 2016: fig. 2), a widespread species in South American savannas, by its cuneate (vs. truncate) leaflets, shorter (c. 10 vs. 16–50 mm long) petiolules, and its 5-angulate (vs. circular) fruits in cross-section.

The characters (leaflet shape, calyx dimensions, and pubescence of staminal tubes) used by Robyns (1963) to segregate Pseudobombax guayasense are not sufficient to distinguish it from P. millei based on a more comprehensive sampling of herbarium collections. Previously, phylogenetic analyses of DNA sequences were interpreted as supporting the accessions examined of these two taxa as sister groups (Carvalho-Sobrinho et al. 2016: fig. 2), but we now consider P. guayasense to be a synonym of P. millei on the basis of combined morphological and molecular evidence.

The characters (tree height and pubescence of staminal tubes) used by Robyns (1967) to segregate P. amapaense do not permit it to be distinguished from P. munguba based on a more comprehensive sampling in the field and in herbarium collections. Previously, phylogenetic analyses of DNA sequences were interpreted as supporting the accessions examined of these two taxa as sister groups (Carvalho-Sobrinho et al. 2016: fig. 2), but we now consider P. amapaense to be a synonym of P. munguba on the basis of combined morphological and molecular evidence.