Pollinator specialisation is one of the major drivers of angiosperm diversification in the Greater Cape Floristic Region (GCFR) of South Africa. Conophytum (Aizoaceae) is a flagship genus for the underexplored arid regions of the GCFR (ca. 108 spp.) with 83.9% of its species endemic to this region and has a floral structure that is unique within the rapidly diversified Ruschieae (Aizoaceae). Floral traits, together with leaf characters divide the genus into 16 sections. We present here the first phylogenetic hypothesis for Conophytum, based on molecular data. The combined phylogenetic data for six plastid regions (matK, rpl16, rps16, trnL-F, trnQ-rps16 and trnS-trnG) were analysed using Bayesian inference, maximum likelihood and maximum parsimony to test the evolution of this floral diversity and current sectional classification. Conophytum was recovered as monophyletic including the three small genera Berrisfordia L.Bolus, Herreanthus Schwantes and Ophthalmophyllum Dinter & Schwantes. Six strongly supported clades were recovered within Conophytum: while several of the sections were broadly retrieved within these clades, the more morphologically variable sections such as Minuscula and Wettsteinia were not supported by our data. The distributions of the six clades largely overlap and are generally confined to the arid parts of the GCFR within South Africa, with only one clade extending into the southern wetter parts. Ancestral character reconstructions showed that floral traits evolved multiple times with low phylogenetic signal recovered for autumn- and winter-flowering and flower type, while the other traits indicated no phylogenetic signal. This lack of phylogenetic signal suggests that drivers of diversity in Conophytum are possibly linked to adaptation to pollinators, with a high rate of inferred pollination shifts observed, i.e. 1.7 shifts per species, mirroring that of Lapeirousia (Iridacaeae) in the GCFR. The evolution of unique tubular flowers and shift to autumn-flowering were recovered as traits indicating strong phylogenetic signal and may have enabled Conophytum to exploit a range of pollinators through diversification in floral morphologies.