Abstract. Dengue-1 virus PDK13 and isolates from vaccinees (dengue-1 Ib1 and dengue-1 Ib 10), dengue-3 PGMK30F3, and dengue-4 PDK48 were studied for their abilities to infect, disseminate, and replicate in Aedes aegypti mosquitoes by the oral route. In general, infection and dissemination rates were poorer for the vaccine compared with the parent viruses. The transmissibility was also lower for dengue-1 PDK13 than the parent virus, whereas it was not detected for isolates from vaccinees and dengue-3 PGMK30F3. Transmissibility of dengue-4 PDK48 was not determined because no dissemination occurred. Replication rates of vaccine strains were also found to be less efficient than the parent viruses. These imply that vaccination with the candidate vaccine is safe. Moreover, vector attenuation of vaccine viruses was consistent with its phenotypic markers of attenuation, which remained stable after a mosquito passage or after human and mosquito passage.Dengue (DEN) is a global disease of the tropics and one of the most important emerging diseases, with half of the world population at risk.1 Since the 1940s, an approach of disease prevention has been directed towards a search for a vaccine.2,3 Live, attenuated, candidate DEN-1 PDK13, DEN-2, PDK53, DEN-3 PGMK30F3, and DEN-4 PDK48 vaccine viruses were developed by the Center for Vaccine Development at Mahidol University (Nakornpathom, Thailand). [4][5][6][7] The DEN-1 PDK13, DEN-2 PDK53, and DEN-4 PDK48 candidate vaccine viruses were derived from serially passages of the parent viruses DEN-1 16007, DEN-2 16681 and DEN-4 1036 in primary dog kidney (PDK) cell cultures (National Institute of Public Health and Environment Protection, Bilthoyen, The Netherlands). 4,8 The DEN-3 PGMK30F3 vaccine was derived from serial passages of DEN-3 16562 virus in primary green monkey kidney (PGMK) cells for 30 consecutive passages and an additional 3 passages in certified fetal rhesus lung (FRhL) cells (American Type Culture Collection, Rockville, MD). These DEN virus candidate vaccines have met the appropriate requirements of biologic attenuation. 4 However, during phase 1 and phase 2 clinical evaluations of DEN candidate vaccines, some volunteers experienced a transient period of viremia, during which dengue viruses were recovered. 5 The possibility of the vector Aedes aegypti feeding on a viremic individual exists, especially when immunization programs are conducted. This concern has provoked an investigation of the stability of the biologic characteristics of DEN candidate vaccines after oral infection into Ae. aegypti. It is expected that the vaccine should not infect mosquito by oral route. However, if it does, the vaccine virus should infect the mosquito poorly and be inefficiently transmitted. Moreover, the vaccine virus should not revert to a virulent form, so that immunization with the vaccine will not disseminate pathogenic dengue virus in nature. Therefore, vector attenuation is an important consideration for a dengue candidate vaccine.Our previous study on DEN-2 PDK53 candidate v...