The proposed phylogeny of Gaura section Gaura is supported by experimental hy- bridization studies. The generally high average capsule-set of crosses between species of section Gaura (78%), compared with capsule-set averages for intersectional crosses (32%), agrees with a pattern of morphological similarity and geographical congruence in confirming the close relation- ships among these species. Higher average crossing success for G. longiflora and its derivatives, G. biennis and G. demareei, suggests the central importance of G. longiflora within the section. Measures of chromosomal diversity demonstrate the apparent conservation of G. longiflora chromosome end arrangements within both G. angustifolia and G. neomexicana, as well as supporting the hypothesis of a hybrid origin for G. demareei (G. lindheimeri x G. longiflora). Apparent levels of translocation diversity in natural populations depend upon the breeding system. Selfing tends to distribute structurally diverse arrangements among homozygous individuals. The extreme chromosomal di- versity of G. biennis sets it apart as a unique genetic system, as does its permanent translocation heterozygosity. Segregation distortion for chromosomal complexes from a ring of four chromo- somes is present in one individual of G. neomexicana. Speciation in section Gaura has not depended upon post-mating, reproductive isolating mechanisms. Factors that contribute to the maintenance of each species as a distinct natural entity are: geographical isolation due to ecological specialization and/or range expansion (G. angustifolia, G. biennis, G. lindheimeri); self-compatibility (G. angustifolia, G. biennis, G. neomexicana); displacement of seasonal (G. angustifolia, G. lindheimeri) and daily (G. demareei, G. lindheimeri) flowering times; and permanent translocation heterozygosity (G. biennis).