The systematics of the Plio-Pleistocene scimitar toothed felid Homotherium have remained problematic after more than a century of fossil findings in Eurasia, Africa and the Americas. Ranging in age between around 4 million and 10,000 years, fossils of this genus display considerable variation, but the distribution of that variation has largely failed to fit a consistent pattern that would allow a clear distinction between species, especially in the Eurasiatic record. The study of undescribed mandibular and cranial fossils of Homotherium from Pleistocene sites in Spain and Alaska provides new insights into the morphological variability within this widespread genus. The results of our study and comparison of the new material with the published fossils of Homotherium confirm the difficulty in dividing the sample into clear-cut species. The new mandible from Incarcal (Spain) shows in a more dramatic way than before how the sample from that Spanish site encompasses the range of variability observed in the Villafranchian and Pleistocene Eurasiatic record, while older, possibly Ruscinian fossils of Homotherium from East Europe display less reduced lower premolars and probably correspond to a different species. The Alaskan fossils, on the other hand, add to the variability in mandibular and cranial morphology of the late Pleistocene North American record. We find no evidence to allow a species-level division within the Villafranchian-Pleistocene Homotherium sample from Eurasia, which for now is best classified as a single variable species, Homotherium latidens, but there are indications of evolution within the lineage, such as the presence of a pocketed anterior margin of the mandibular masseteric fossa, a feature found in the younger fossils of middle or late Pleistocene age but consistently absent in older specimens. A comparable pattern is found in the American record, where the same mandibular feature is observed in late Pleistocene fossils, although in that continent the “primitive” features of some older Homotherium fossils of Blancan (Pliocene) age are consistent enough to justify their classification in a separate species, Homotherium ischyrus. Only the finding of more complete cranial fossils of middle and late Pleistocene age will reveal if there are additional morphological features (besides the pocketing of the masseteric fossa) that could one day allow the separation of younger populations from those of Late Pliocene and Early Pleistocene age.