All 100+ bedbug species (Cimicidae) are obligate blood-sucking parasites [1, 2]. In general, blood sucking (hematophagy) is thought to have evolved in generalist feeders adventitiously taking blood meals [3, 4], but those cimicid taxa currently considered ancestral are putative host specialists [1, 5]. Bats are believed to be the ancestral hosts of cimicids [1], but a cimicid fossil [6] predates the oldest known bat fossil [7] by >30 million years (Ma). The bedbugs that parasitize humans [1, 8] are host generalists, so their evolution from specialist ancestors is incompatible with the "resource efficiency" hypothesis and only partially consistent with the "oscillation" hypothesis [9-16]. Because quantifying host shift frequencies of hematophagous specialists and generalists may help to predict host associations when vertebrate ranges expand by climate change [17], livestock, and pet trade in general and because of the previously proposed role of human pre-history in parasite speciation [18-20], we constructed a fossil-dated, molecular phylogeny of the Cimicidae. This phylogeny places ancestral Cimicidae to 115 mya as hematophagous specialists with lineages that later frequently populated bat and bird lineages. We also found that the clades, including the two major current urban pests, Cimex lectularius and C. hemipterus, separated 47 mya, rejecting the notion that the evolutionary trajectories of Homo caused their divergence [18-21]. VIDEO ABSTRACT.