The main theoretical issue of this paper is the interpretation of the close similarity of the total behaviour and colour patterns of 3 not closely related Barbus species. The case for evolutionary convergence is easy if the total pattern of each species is interpreted as a collection of evolutionary adaptations to the species' natural habitat. The thorough similarity of the 3 species must then be ascribed to key similarities between their natural habitats. However, it is argued that another interpretation is possible, in which the diverse traits cohere causally and have been acted upon as one whole by natural selection. In the extreme case of this alternative interpretation, the case for convergence may still be won, but it is a Pyrrhic victory: convergence has lost much of its usual meaning. Apart from the above discussion, the reproductive behaviour and spawning colours of the 3 species are described and compared in some detail. Possible functional reasons for a relatively small difference as to the degree of sexual colour dimorphism are discussed in relation to eye sizes and visual discrimination.