Leaf anatomical and chemotaxonomic studies were carried out on the subtribe Gnaphaliinae (Inuleae - Compositae) with the aim of clarifying the status and relationships of the New Zealand and Tasmanian taxa. The leaf anatomy and flavonoid patterns of 48 species were compared. Data were analysed numerically and cladistically. The Gnaphaliinae is a problematical group, especially in terms of generic boundaries. In the leaf anatomical and chemotaxonomic analyses, the New Zealand species of AnaphaJis are similar to each other but quite different from a Himalayan representative (Anaphalis triplinervis). In the New Zealand species of Hel/chrysum the relationships of the two species of section Xerochlaena are unresolved but there Is evidence that the New Zealand species of section Ozothamnus should be treated separately from HeJichrysum. The New Zealand species of Cassinia are very similar to the Tasmanian species of Helichrysum sect. Ozothamnus but not to the Tasmanian species of Cassinia. Ewartia sinclairil of New Zealand is very different from the Tasmanian species of Ewartia. The exclusion of the anaphalioid species of GnaphaJium and of GnaphaJium lutee-album from GnaphaJium Is supported. The species of GnaphaJium sect. Euchiton are not a coherent group. It appears that RaouJia should be split into two genera but with a few species excluded from both segregates. Genus HZ: has close links to neither Leucogenes nor Haastia. Its separate generic status seems to be justified. The species of Leucogenes are closely related. Haastia and Pterygopappus are not close to each other or to any other genus. It is shown that leaf anatomical and flavonoid data provide taxonomically useful characters for the classification of the Gnaphaliinae.