Vegetative propagation (clonal growth) conveys several evolutionary advantages that positively affect life history fitness and is a widespread phenomenon among angiosperms that also reproduce sexually. However, a bias towards clonality can interfere with sexual reproduction and lead to sexual extinction, although a dearth of effective genetic tools and mathematical models for clonal plants has hampered assessment of these impacts. Using the endangered tropical epiphytic or lithophytic orchid Bulbophyllum bicolor as a model, we integrated an examination of breeding system with 12 microsatellite loci and models valid for clonal species to test for the ‘loss of sex’ and infer likely consequences for long-term reproductive dynamics. Bagging experiments and field observations revealed B. bicolor to be self-incompatible and pollinator-dependent, with an absence of fruit-set over four years. Challenging the assumptions that clonal populations can be as genotypically diverse as sexually reproducing ones and that clonality does not greatly influence genetic structure, just 22 multilocus genotypes were confirmed among all 15 extant natural populations, 12 of the populations were found to be monoclonal and all three multiclonal ones exhibited a distinct phalanx clonal architecture. Our results suggest that all B. bicolor populations depend overwhelmingly on clonal growth for persistence, with a concomitant loss of sex due to an absence of pollinators and a lack of mating opportunities at virtually all sites, both of which are further entrenched by habitat fragmentation. Such cryptic life history impacts, potentially contributing to extinction debt, could be widespread among similarly fragmented, outcrossing tropical epiphytes, demanding urgent conservation attention.