Koinophilia is an evolutionary hypothesis proposing that during sexual selection, animals preferentially seek mates with a minimum of unusual or mutant features, including functionality, appearance and behavior. Koinophilia intends to explain the clustering of sexual organisms into species and other issues described by Darwin's dilemma. The term derives from the Greek word koinos meaning "common" or "that which is shared", and philia, meaning "fondness".

Natural selection causes beneficial inherited features to become more common at the expense of their disadvantageous counterparts. The koinophilia hypothesis proposes that a sexually reproducing animal would therefore be expected to avoid individuals with rare or unusual features, and to prefer to mate with individuals displaying a predominance of common or average features. Mutants with peculiar features would be avoided because most mutations that manifest themselves as changes in appearance, functionality or behavior are disadvantageous. Because it is impossible to judge whether a new mutation is beneficial (or might be advantageous in the unforeseeable future) or not, koinophilic animals avoid them all, at the cost of avoiding the very occasional potentially beneficial mutation. Thus, koinophilia, although not infallible in its ability to distinguish fit from unfit mates, is a good strategy when choosing a mate. A koinophilic choice ensures that offspring are likely to inherit a suite of features and attributes that have served all the members of the species well in the past.

Koinophilia differs from the "like prefers like" mating pattern of assortative mating. If like preferred like, leucistic animals (such as white peacocks) would be sexually attracted to one another, and a leucistic subspecies would come into being. Koinophilia predicts that this is unlikely because leucistic animals are attracted to the average in the same way as are all the other members of its species. Since non-leucistic animals are not attracted by leucism, few leucistic individuals find mates, and leucistic lineages will rarely form.

Koinophilia provides simple explanations for the almost universal canalization of sexual creatures into species, the rarity of transitional forms between species (between both extant and fossil species), evolutionary stasis, punctuated equilibria, and the evolution of cooperation. Koinophilia might also contribute to the maintenance of sexual reproduction, preventing its reversion to the much simpler asexual form of reproduction.

The koinophilia hypothesis is supported by the findings of Judith Langlois and her co-workers. They found that the average of two human faces was more attractive than either of the faces from which that average was derived. The more faces (of the same gender and age) that were used in the averaging process the more attractive and appealing the average face became. This work into averageness supports koinophilia as an explanation of what constitutes a beautiful face.

Biologists from Darwin onwards have puzzled over how evolution produces species whose adult members look extraordinarily alike, and distinctively different from the members of other species. Lions and leopards are, for instance, both large carnivores that inhabit the same general environment, and hunt much the same prey, but look quite different. The question is why intermediates do not exist.

This is the "horizontal" dimension of a two-dimensional problem, referring to the almost complete absence of transitional or intermediate forms between present-day species (e.g. between lions, leopards, and cheetahs).

The "vertical" dimension concerns the fossil record. Fossil species are frequently remarkably stable over extremely long periods of geological time, despite continental drift, major climate changes, and mass extinctions. When a change in form occurs, it tends to be abrupt in geological terms, again producing phenotypic gaps (i.e. an absence of intermediate forms), but now between successive species, which then often co-exist for long periods of time. Thus the fossil record suggests that evolution occurs in bursts, interspersed by long periods of evolutionary stagnation in so-called punctuated equilibria. Why this is so has been an evolutionary enigma ever since Darwin first recognized the problem.