Macroevolution in the modern sense is evolution that is guided by selection among interspecific variation, as opposed to selection among intraspecific variation in microevolution. This modern definition differs from the original concept, which referred macroevolution to the evolution of taxa above the species level (genera, families, orders etc.).
Philiptschenko distinguished between microevolution and macroevolution because he rejected natural selection in the sense of Darwin as an explanation for larger evolutionary transitions that give rise to taxa above the species level in the Linnean taxonomy. Accordingly, he restricted Darwinian "microevolution" to evolutionary changes within the boundary of given species that may lead to different races or subspecies at the most. By contrast, he referred "macroevolution" to major evolutionary changes that correspond to taxonomic differences above the species level, which in his opinion would require evolutionary processes different from natural selection. An explanatory model for macroevolution in this sense was the "hopeful monster" concept of geneticist Richard Goldschmidt, who suggested saltational evolutionary changes either due to mutations that affect the rates of developmental processes or due to alterations in the chromosomal pattern. Particularly the latter idea was widely rejected by the modern synthesis and is disproved today, but the hopeful monster concept based on evo-devo explanations found a moderate revival in recent times. As an alternative to saltational evolution, Dobzhansky  suggested that the difference between macroevolution and microevolution reflects essentially a difference in time-scales, and that macroevolutionary changes were simply the sum of microevolutionary changes over geologic time. This view became broadly accepted, and accordingly, the term macroevolution has been used widely as a neutral label for the study of evolutionary changes that take place over a very large time-scale. However, the tenet that large-scale evolutionary patterns were ultimately reducible to microevolution has been challenged by the concept of species selection, which suggests that selection among species is a major evolutionary factor that is independent from and complementary to selection among organisms. Accordingly, the level of selection (or, more generally, of sorting) has become the conceptual basis of a third definition, which defines macroevolution as evolution through selection among interspecific variation.
According to the modern definition, the evolutionary transition from the ancestral to the daughter species is microevolutionary, because it results from selection (or, more generally, sorting) among varying organisms. However, speciation has also a macroevolutionary aspect, because it produces the interspecific variation species selection operates on. Another macroevolutionary aspect of speciation is the rate at which it successfully occurs, analogous to reproduction success in microevolution.
"Species selection operates on variation provided by the largely random process of speciation and favors species that speciate at high rates or survive for long periods and therefore tend to leave many daughter species." Species selection comprises (a) effect-macroevolution, where organism-level traits (aggregate traits) affect speciation and extinction rates (Stanley's original concept), and (b) strict-sense species selection, where species-level traits (e.g. geographical range) affect speciation and extinction rates. It has been argued that effect macroevolution is reducible to microevolution because both operate through selection on organismic traits, but Grantham demonstrated that effect macroevolution can oppose selection at the organismic level and is therefore not reducible microevolution. Cases in which selection on the same trait has opposing effects at the organismic and the species level have been made in the context of sexual selection, which increases individual fitness but may also increase the extinction risk of the species.
Punctuated equilibrium postulates that evolutionary change is concentrated during a geologically short speciation phase, which is followed by evolutionary stasis that persists until the species goes extinct. The prevalence of evolutionary stasis through most of the existence time of species is a major argument for the relevance of species selection in shaping the evolutionary history of clades. However, punctuated equilibrium is neither a macroevolutionary model of speciation, nor is it a prerequisite for species selection.
A macroevolutionary benchmark study is Sepkoski's work on marine animal diversity through the Phanerozoic. His iconic diagram of the numbers of marine families from the Cambrian to the Recent illustrates the successive expansion and dwindling of three "evolutionary faunas" that were characterized by differences in origination rates and carrying capacities.
The macroevolutionary relevance of environmental changes is most obvious in the case of global mass extinction events. Such events are usually due to massive disturbances of the non-biotic environment that occur too fast for a microevolutionary response through adaptive change. Mass extinctions therefore act nearly excursively through selection among species, i.e., macroevolutionary. In their differential impact on species, mass extinctions introduce a strong non-adaptive aspect to evolution. A classic example in this context is the suggestion that the decline of brachiopods that is apparently mirrored by the rise of bivalves was actually caused by differential survival of these clades during the end-Permian mass extinction.
Macroevolution is driven by differences between species in origination and extinction rates. Remarkably, these two factors are generally positively correlated: taxa that have typically high diversification rates have also high extinction rates. This observation has been described first by Steven Stanley, who attributed it to a variety of ecological factors. Yet, a positive correlation of origination and extinction rates is also a prediction of the Red Queen hypothesis, which postulates that evolutionary progress (increase in fitness) of any given species causes a decrease in fitness of other species, ultimately driving to extinction those species that do not adapt rapidly enough. High rates of origination must therefore correlate with high rates of extinction. Stanley's rule, which applies to almost all taxa and geologic ages, is therefore a strong indication for a dominant role of biotic interactions in macroevolution.
Subjects studied within macroevolution include: