The Rift Valley fever virus (RVFV) is a threat that must not be neglected, as the consequences of RVFV are dramatic, both for human and animal health. This virus is a zoonotic virus that already has demonstrated a real capacity for re-emerging after long periods of silence, as observed in Barkedji (Senegal, West Africa) in 2002. In this article we present the 2nd emergence in Barkedji after the 1st manifestation in 1993, and for the 1st time the circulation of RVFV during 2 consecutive years among mosquito populations in Senegal. As part of the entomological surveillance program undertaken since 1990 to detect circulation of the RVFV in Barkedji, 108,336 mosquitoes belonging to 34 species and 5 genera were collected in 2002–2003. Aedes vexans and Culex poicilipes, previously known to be vectors of RVFV in Senegal, comprised 88.7% of the total collection. In 2002, Ae. vexans was the most abundant mosquito, followed by Cx. poicilipes; the opposite situation was observed in 2003. In 2002, 29 and 10 RVFV isolates were obtained from Cx. poicilipes (minimum infection rate [MIR]  =  0.13%) and Ae. vexans (MIR  =  0.02%) pools, respectively and the MIR for the 2 species were significantly different (χ²  =  34.65; df  =  1, P < 0.001). In 2003, 7 RVFV strains were isolated from Cx. poicilipes (3, MIR  =  0.03), Mansonia africana (2, MIR  =  0.08), Ae. fowleri (1), and Ma. uniformis (1, MIR  =  0.05). The 3 latter species were found to be associated with RVFV for the 1st time in Senegal. A significant decrease in MIR was observed from 2002 to 2003 (χ²  =  6.28; df  =  1, P  =  0.01) for Cx. poicilipes, the only species involved in the transmission during the 2 sampling years.