Traps containing a combination floral and synthetic pheromone lure are used to monitor and manage Japanese beetles, Popillia japonica Newman (Coleoptera: Scarabaeidae). One key factor limiting trap effectiveness for beetle control is the “trap spillover” phenomenon, in which the trap attracts beetles without capturing them, resulting in increased damage to surrounding host plants. We investigated the mechanisms underlying trap spillover by conducting two studies in a soybean field in east central Illinois. In the first study, we set up trap stations for 1 d and compared the sex, size, and egg load (for females) of beetles caught in the traps with those on the plants immediately surrounding the trap, downwind of the trap, at lure-only (no trap) stations, and at control areas. Females caught in traps tended to be smaller than those on plants surrounding the traps, and females attracted to the traps had fewer eggs than those downwind or at control sites. We did not find any difference in male characteristics. In the second study, we observed the behavior of beetles initially approaching traps. Upon initial approach, the majority of individuals landed on plants before making contact with the trap, and those beetles that spent an extended time on the leaves tended to be females. Arriving males would occasionally pair with these females on the plants. Overall, traps did not capture a random subset of the beetles present in the field. We hypothesize that trap spillover is a result of arriving females not being as attracted to the precise location of the trap as they are to the general location itself, and of arriving males seeking mates and finding them among these spillover females.