Attack rates of the parasitoid Leptomastix dactylopii (Howard) were examined in petri dishes and on caged plants that varied in architectural characteristics. Individual female parasitoids were placed into petri dishes containing a range of densities of the citrus mealybug, Planococcus citri (Risso). Green and yellow-variegated coleus, Solenostemon scutellarioides (L.) Codd, were used to determine the effects of plant architecture on the rate at which L. dactylopii parasitized citrus mealybugs. Coleus plants were categorized into size classes based on height, number of leaves, leaf surface area, and number of branches. A single mated female L. dactylopii was placed into each plastic observation cage containing a plant with 1, 8, or 16 third- to early fourth-instar mealybugs and allowed to forage for 24 h. After this time, citrus mealybugs were removed from coleus plants and placed into petri dishes that were checked after 10 d for mummified citrus mealybugs, the measure of a successful attack. L. dactylopii were unable to find a host after 24 h when only one citrus mealybug was present in the searching environment. Higher attack rates occurred as the number of mealybugs increased in the petri dishes. L. dactylopii attacked an average of 15.6 ± 2.3 mealybugs within a 24-h period. Higher attack rates were evident as the number of citrus mealybugs increased on caged plants. Plant color had no effect on the attack rate of L. dactylopii. All architectural characterizations of plant size, height, leaf number, leaf surface area, and branch number were negatively correlated with parasitoid attack rate. These findings suggest that biological control practitioners may use any convenient measure of plant size (e.g., height) to modify the release rate of L. dactylopii in a citrus mealybug management program.