This study evaluated azadirachtin and imidacloprid for their ability to reduce injury by Japanese beetles [Popillia japonica (Coleoptera: Scarabaeidae)] on floribunda-type roses (Rosa sp. ‘Acadia Sunrise’), either applied to foliage or as a soil drench. Roses were arranged in field plots and exposed to resident adult beetle populations. Insecticides were evaluated in field and laboratory trials. Laboratory assays of leaves collected from plants 14 days after soil applications of azadirachtin were less preferred by adult beetles than those collected from untreated controls. Plants in field trials that received soil treatments of either imidacloprid or azadirachtin had defoliation levels that were <8% throughout the entire season, whereas untreated control plants were 20% defoliated. Addition of foliar sprays to soil applied insecticides provided no added protection to foliage. Rose blooms were more difficult to protect with both foliar and soil-applied insecticides. Bloom injury of untreated controls varied between 20% and 30%, while plants receiving soil applications of azadirachtin varied between 0.2% and 18%. Soil applications of imidacloprid provided somewhat better protection of blooms with injury ranging between 0.2% and 8%. Foliar applications of azadirachtin gave no added protection to blooms of plants treated with imidacloprid. Adding carbaryl foliar treatments every 2 weeks improved control to <2% injury, a level that was comparable to weekly application of carbaryl. The potential for using soil-applied azadirachtin to reduce the need for foliar applications of carbaryl in rose gardens is discussed.