Most modern carnation (Dianthus caryophyllus L.) cultivars have weak fragrances dominated by the scent of methyl benzoate. Wild Dianthus species with strong or unique scents may be useful gene resources for the improvement of carnation fragrances. We investigated the scents of interspecific hybrids between carnations and fragrant wild species by gas chromatography-mass spectrometry (GC-MS), and evaluated the usefulness of wild species for fragrant breeding in carnations. Dianthus hungaricus, which produced large amounts of various benzenoids, was crossed with a carnation with a floral scent dominated by methyl benzoate, but benzenoid diversity was not increased in the interspecific hybrid. We also analyzed some existing interspecific hybrids. Dianthus superbus var. longicalycinus had high amounts of β-ocimene and β-caryophyllene. These terpenoids were acquired as principal scent compounds by some interspecific hybrids between this species and a carnation lacking terpenoids. Three unidentified wild species (Dianthus sp. 4, 5, and 6) emitted high amounts of benzenoids, including eugenol, benzyl alcohol, methyl o-anisate, and methyl salicylate. These benzenoids were also detected in interspecific hybrids between carnations and the wild species, and the amounts were increased compared to the parental carnation. The emission of these scent compounds of wild Dianthus species was inherited by most hybrids lines; the variety and amounts of scent compounds tended to increase compared to parental carnations, although there was no general hereditary pattern. As we actually sensed the fragrances of the principal compounds from some hybrid flowers, the usefulness of interspecific hybridizations for the improvement of flower fragrances was confirmed. Dianthus superbus var. longicalycinus and Dianthus sp. 4, 5, and 6 seemed promising resources regarding the addition of terpenoids and the increase in benzenoid variation in the floral volatiles of carnations.