Odors were collected from the headspace above fermented infusions of Bermuda grass, which commonly are used as attractants in traps for gravid mosquitoes. Coupled gas chromatography-electroantennogram detection (GC-EAD) was used to identify 9 compounds (phenol, p-cresol, 4-ethylphenol, indole, 3-methylindole, nonanal, 2-undecanone, 2-tridecanone, naphthalene) from odor extracts that elicited significant antennal responses from antennae of gravid female Culex quinquefasciatus Say and Culex tarsalis Coquillett mosquitoes. Several of these compounds at appropriate concentrations were weakly attractive to gravid female mosquitoes in laboratory bioassays and/or stimulated more oviposition than water controls. In addition, dimethyltrisulfide, a significant component of odor extract which did not elicit strong responses from female mosquito antennae in GC-EAD assays, also appeared to stimulate oviposition at 1 concentration. A reconstituted blend of the 10 compounds resulted in much stronger and more consistent responses than individual compounds. The blend was attractive to gravid Cx. quinquefasciatus and enhanced oviposition over a range of concentrations spanning 3 orders of magnitude. One concentration of the blend also attracted gravid Cx. tarsalis and enhanced oviposition. However, at the highest concentration tested, the blend was repellent to both species. Overall, these studies demonstrated that gravid mosquitoes are attracted to oviposition sites by blends of compounds rather than by individual chemicals, and that the concentration of compounds in the odor is a critical factor in determining whether responses are positive (attractive, stimulatory) or negative (repellent, deterrent).