The estimation of hybrid frequency in a flora is fraught with problems. It requires a taxonomy wherein: (a) putative parental orthospecies are correctly delimited; (b) nothospecies are not mistaken for orthospecies; (c) the flora is sufficiently well known that there is confidence that most putative hybrids have been discovered. The depth of historic collecting in the state, the interest in Missouri Crataegus shown by experts in the genus, together with the large number of taxa described from the state, collectively make Missouri an excellent candidate for analysis. My data indicate that 20% of possible hybrid combinations have occurred over the last 120 years. The great majority of putative hybrids appear to have had small evanescent populations and are now apparently extinct. A few nothospecies, however, survive today in scattered, small populations, perhaps by being repeatedly created. Interestingly, extrapolation (admittedly with less robust data) to the whole of the northeastern United States and adjacent Canada, provides a similar figure, 22%. There is also informal support for an interpretation that has hybrid occurrence more likely (given sympatry and similar anthesis phenology) between closely related than more distantly related Crataegus species. If these results are shown to hold up with molecular analysis then it is clear that well over 99% of individuals encountered in the field belong to orthospecies. These insights suggest that a radical reappraisal is due of the widely held view that hybridization is one of the chief problems in North American Crataegus taxonomy.

Key words: Crataegus, hybridization, Missouri, North America, Rosaceae