Putative gamete-derived progenies from two Hylocereus species, the diploid H. monacanthus and the tetraploid H. megalanthus, were studied with the dual aims to confirm their gamete origin and to evaluate their potential use as genetic resources. An additional goal was to determine the origin (allotetraploid vs. autotetraploid) of H. megalanthus by exploring morphological variations in the di-haploid (2x) H. megalanthus progeny. Gamete origin was proved in all five H. monacanthus lines obtained and in 49 of the 70 H. megalanthus lines by using flow cytometry and simple sequence repeat (SSR) markers. The five double-haploid (2x) H. monacanthus lines showed low vigor and abnormal flower development, with malformed ovules and aborted pollen grains. Only one flower set fruit, giving several viable seeds. For H. megalanthus, both abnormal ovules and defective anthers were observed in the di-haploid (2x) and double di-haploid (4x) lines. Among the 46 di-haploid lines, only 14 set fruit. Another 13 di-haploid lines formed flower buds that abscised before anthesis or soon after pollination. The severe sterility of the double-haploid H. monacanthus and the reduced fertility of all the di-haploid and double di-haploid H. megalanthus lines can be linked to their reduced heterozygosity, which drastically affected the development of normal female and male organs. We thus concluded that chromosome doubling, as occurred spontaneously in the double-haploid H. monacanthus and the double di-haploid H. megalanthus, is not sufficient to restore fertility in Hylocereus. We also observed very low gametoclonal variation among the di-haploid (2x) H. megalanthus lines, a finding that supported an autotetraploid, rather than an allotetraploid, origin of this species. Nonetheless, despite the above-described challenging limitations, these gamete-derived lines are currently being bred as the seed parent, offering unique possibilities for genetic research and additional breeding.