The forces driving regression of biologically functionless traits remain disputed. There is ongoing debate regarding whether selection, as opposed to disuse and neutral mutations, is involved in this process. Cave species are of particular relevance for study in this regard because in continuous darkness all traits that depend on information from light, such as eyes, dark pigmentation and certain behaviours, abruptly lose their function. Recently, strong selection driving reduction has again been proposed, which relied on modelling analyses based on assumptions such as immigration of surface alleles into the cave forms or no fitness difference existing between Astyanax surface and cave fish. The validity of these assumptions, often applied to reject neutral processes in functionless traits, is questioned in this review. Morphological variation in a trait resulting from genetic variability is typical of biologically functionless traits and is particularly notable in phylogenetically young cave species. It is the most evident indicator of loss of selection, which normally enforces uniformity to guarantee optimal functionality. Phenotypic and genotypic variability in Astyanax cave fish eyes does not derive from genetic introgression by the surface form, but from regressive mutations not being eliminated by selection. This matches well with the principles of Kimura’s neutral theory of molecular evolution.