Müllerian mimicry, in which there is convergence in coloration between unrelated unpalatable species, should lead to uniformity in appearance, not polymorphism, and so the occurrence in tropical Africa of unrelated species of unpalatable butterflies with corresponding polymorphic colour forms suggested a problem of special evolutionary interest. Field work in Uganda and Sierra Leone in 1964 72 demonstrated a statistical association between the occurrence and relative frequencies of polymorphic forms inDanaus chrysippus(Danainae) andAcraea encedon(Acraeinae) which was deemed as confirmation of a Müllerian relationship between them. There were, however, certain anomalies which at the time remained unresolved. Later, in 1976, it was discovered that what had been calledA. encedonis in reality two sibling species.A. encedonand a new one, named asA. encedana.The two differ in the structure of both male and female genitalia and in the coloration and the food‐plants of the larvae. The recognition of the additional species has enabled a re‐assessment of the polymorphie Müllerian association withD. chrysippus.It emerges that, although there is a close qualitative and quantitative reciprocal mimetic relationship betweenA. encedanaandD. chrysippus,the relationship betweenA. encedonandD. chrysippusis much weaker, and in places non‐exitent. The possible origin of the mimetic polymorphism is discussed in terms of hybridization of previously allopatric and monomorphic populations which have met as a consequence of recent expansions of geographical range in all three species resulting from forest clearance and the spread of savanna‐like conditions in previously fore