Relationships within Loxia are muddy chiefly because species limits within the L. curvirostra complex are controversial. That aside, it is clear that there are two species complexes in the genus: the L. curvirostra complex, which contains between one and ten (or more) species, and the L. leucoptera complex, which contains between one and three species, in the latter case with species limits corresponding to named subspecies. Loxia megaplaga Riley, 1916, the Hispaniolan Crossbill, was considered a subspecies of L. leucoptera for many decades, but on the basis of vocal, morphological, and genetic divergence (Benkman 1994, Smith 1997, Parchman et al. 2006, 2007), the taxon generally is treated as a species now (e.g., Banks et al. 2003). Loxia megaplaga diverged from L. leucoptera roughly 550,000 years ago (Parchman et al. 2007). Given marked differences in voice (Elmberg 1993), coupled with morphological (and posited ecological) differences, Old World and New World L. leucoptera also may be species rather than subspecies.
Hybrids between species groups are unknown in the wild. A male L. leucoptera caged with a female L. curvirostra produced two broods of three young each (H. B. Tordoff pers. comm.). A resultant male hybrid, which had wing bars and bill size intermediate to its parents, was backcrossed with a female L. curvirostra. The offspring "appeared essentially like" L. curvirostra (H. B. Tordoff pers. comm.).