A hybrid zone exists where the ranges of two interbreeding species or diverged intraspecific lineages meet and cross-fertilize. Hybrid zones can form in situ due to the evolution of a new lineage^{[page needed]} but generally they result from secondary contact of the parental forms after a period of geographic isolation, which allowed their differentiation. Hybrid zones are useful in studying the genetics of speciation as they can provide natural examples of differentiation and gene flow between populations that are at some point on the continuum between diverging populations and separate species with reproductive isolation.

Hybrid zones are areas where the hybrid offspring of two divergent taxa (species, subspecies or genetic "forms") are prevalent and there is a cline in the genetic composition of populations from one taxon to the other. The two (or more) genetically differentiated species or lineages contributing to formation of a hybrid zone are regarded as parental forms. Precise definitions of hybrid zones vary; some insist on increased variability of fitness within the zone, others that hybrids be identifiably different from parental forms and others that they represent secondary contact alone. The widths of such zones can vary from tens of metres to hundreds of kilometres. The shape of the zones (clines) can be gradual or stepped. Additionally, hybrid zones may be ephemeral or long-lasting.

Some hybrid zones can be seen as presenting a paradox for the biological definition of a species, usually given as "a population of actually or potentially interbreeding individuals that produce fertile offspring" under what has become known as the Biological Species Concept. Under this definition, both parental forms could be argued to be the same species if they produce fertile offspring at least some of the time. However, the two parental populations or species often remain identifiably distinct, conforming to an alternative, and presently preferred concept of species as "taxa that retain their identity despite gene flow".

The clines of hybrid zones can be observed by recording the frequency of certain diagnostic alleles or phenotypic characteristics for either population along a transect between the two parental populations or species. Often the clines take the form of a sigmoidal curve. They can be wide (gradual) or narrow (steep) depending on the ratio of hybrid survival to recombination of genes. Hybrid zones which show no regular transition from one taxon to the other, but rather a patchy distribution of parental forms and subpopulations with hybrid background, are termed mosaic hybrid zones.

Various models and theories have been developed by the researchers of hybrid zones. Major models can be largely categorized into four types: ephemeral hybrid zone theory, bounded hybrid superiority model, mosaic hybrid zone model and tension zone model. In each model, different evolutionary forces are attributed different levels of importance. The different models result largely from the study of different biological material (natural populations). The four major models operate mostly under a general framework of either a balance between natural selection and dispersal or interaction between genotypes and environment. Different hybrid zones may fit different models and no single theory or model serves to explain all the hybrid zones found in nature.

Some early evolutionary biologists who preferred a biological species concept, such as Ernst Mayr and Theodosius Dobzhansky, believed that hybrid zones are generally rare and ephemeral, with an eventual fate of either merging of the hybridizing populations or reinforcement, which leads to a speciation event. The extinction of one of the hybridizing populations through introgression is sometimes termed "waves of advance". (Although this term can also refer to the spreading of advantageous allele across a reproductive barrier) The ephemerality of hybrid zone has been countered by the discovery of many hybrid zones that has lasted for a long period of time, up to 100,000 years found between the iguanid lizards, Sceloporus woodi and S. undulatus undulatus.

The bounded hybrid superiority model predicts that hybrids have higher fitness in a habitat that is intermediate between those of their parental populations. The hybrid habitat occurs usually, but not necessarily, on a narrow ecotone. Clines of a bounded hybrid superiority zone reflect a smooth gradient corresponding to the gradient of differential selection strength in fitness-related characteristics. The bounded superiority model places a high importance on the ecological aspect of the habitat. In fact, botanist Edgar Anderson suggested that hybrid populations are more likely to inhabit ecologically disturbed areas, which often occur under human’s modification of landscapes or geological events that create novel habitat conditions. He argued that hybrid zones are essentially formed via "hybridization of habitats". Anderson also considered natural hybridization as a positive evolutionary stimulus that allows different populations and lineages to exchange adaptive genetic elements—similar view that place a high evolutionary importance on hybrid zones is more prevalent among botanists, in contrast to zoologists who are more likely see hybrid zones as more of a "natural laboratory" of population genetics. Some criticism for the bounded superiority model suggests that hybrid zones with higher hybrid fitness are theoretically unlikely to be distributed along narrow ecotones; some also point out that there has not been direct empirical evidence of higher hybrid fitness along an ecological gradient (i.e. ecotones).

The term tension zone was first used by K. H. L. Key to describe an area of hybridizing populations that act like a "semipermeable membrane" in terms of gene exchange. This term was later taken by Nicholas Barton and Godfrey Hewitt to denote a hybrid zone maintained by a balance between selection and dispersal. Similar models have also previously been termed dynamic equilibrium.