Chagas disease affects millions of people in Latin America. The control of this vector-borne disease focuses on halting transmission by reducing or eliminating insect vector populations. Most transmission of Trypanosoma cruzi, the causative agent of Chagas disease, involves insects living within or very close to households and feeding mostly on domestic animals. As animal hosts can be intermittently present it is important to understand how host availability can modify transmission risk to humans and to characterize the host-seeking dispersal of triatomine vectors on a very fine scale. We used a semi-field system with motion-detection cameras to characterize the dispersal of Triatoma infestans, and compare the behavior of vector populations in the constant presence of hosts (guinea pigs), and after the removal of the hosts. The emigration rate - net insect population decline in original refuge - following host removal was on average 19.7% of insects per 10 days compared to 10.2% in constant host populations (p = 0.029). However, dispersal of T. infestans occurred in both directions, towards and away from the initial location of the hosts. The majority of insects that moved towards the original location of guinea pigs remained there for 4 weeks. Oviposition and mortality were observed and analyzed in the context of insect dispersal, but only mortality was higher in the group where animal hosts were removed (p-value <0.01). We discuss different survival strategies associated with the observed behavior and its implications for vector control. Removing domestic animals in infested areas increases vector dispersal from the first day of host removal. The implications of these patterns of vector dispersal in a field setting are not yet known but could result in movement towards human rooms.