The purpose of the study was to determine direction of mosquito dispersal for species Culex tarsalis and to examine the environmental effect on dispersion. Culex tarsalis Coquillett is the most abundant mosquito species in southeast California and the primary rural vector of the area. It transmits a wide variety of viruses; the most outstanding are Western equine encephalomyelitis (WEE), St. Louis encephalitis (SLE) and West Nile virus (WNV). The investigation of dispersal of Cx. tarsalis was performed in the area of Mecca town in Coachella Valley, California, in the period October 22 – November 6, 2007, using capturemark-release-recapture methods. First, the mosquitoes (23,890) were collected at traps placed south from Mecca, then marked with fluorescent dust (orange and aqua) and, finally, released at two locations: northeast and southwest of small town Mecca. Recapture was performed nine times in twelve days at 40 CO2 baited CDC-traps without lights and placed across the area (30.6 km2) in six different habitats (desert, date gardens,citrus orchards, vineyards, agricultural area and urban area). The sampling points were between 0.2 km to 0.8 km away from each other. During the first three days of investigation 90.22% of the total 266 Cx. tarsalis females (1.11%) were recaptured. More than 50% of the marked mosquitoes were collected in the desert while only 3.4% in urban habitat. The analysis indicated that in the surroundings of the trap site the number of marked mosquitoes significantly increases along with the raise in the portion of the habitat characterised by the desert and significantly decreases with the raise in the portion of the habitat that is urban. However, the portion of agricultural area, citrus orchards, vineyards and date gardens has no statistically significant bearing. Statistical analyses confirmed that there were no differences in dispersion between orange and aqua mosquitoes (dependent on the distance from the release point and on the angle of deflection from the North). A significant tendency of mosquito dispersal to the south was recorded for both release points, probably due to the most favourable feeding places and oviposition sites and potentially increase humidity caused by the vicinity of large ponds. Furthermore, there is a possibility that the wind, predominantly northern during the investigation period, had an influence on mosquito dispersal direction. The mean distance travelled was different for aqua (0.79 km) and orange (1.06 km) mosquitoes. Eight species emerged through faunistic analysis of the adult mosquitoes captured (34,708), the most common being Cx. tarsalis (86.09%). This is the only eudominant, euconstant and exclusive species in all habitats. It was determined that mosquito population structure and mosquito abundance oscillate in accordance with the habitat type. The comparison of dispersion of rural and floodwater mosquito species identified differences in dispersal rates. However, similarities were detected with regard to the direction (the prevailing direction was away from urban habitats). Knowledge of mosquito movement obtained by the investigations in California and Eastern Croatia can improve mosquito and vector control programs by taking into account environmental factors that influence female mosquito flight direction.