Hexazinone (1.12 kg ai/ha), triclopyr (1.12 kg ai/ha), metsulfuron (0.008 kg ai/ha), dichlorprop + 2,4-D, glyphosate (2.8%), and triclopyr (2%) + diesel oil (98%), applied as a broadcast or spot (individual plant) treatment, were evaluated over two years in south Florida for tropical soda apple (TSA) control and their effects on grass ground cover. For broadcast treatments, triclopyr (98%) and hexazinone (93%), had significantly (P < 0.05) higher percent control of marked TSA plants 90 d after herbicide application. However, triclopyr (99%) had significantly higher grass ground cover than hexazinone (78%). Hexazinone severely damaged Pangola digitgrass, but had no effect on bahiagrass. For spot treatments, dichlorprop + 2,4-D (100%) had the highest percent total control of TSA and least effect on grass ground cover (96%) 90 d after herbicide application, followed by glyphosate (96% control) and triclopyr + diesel oil (95% control). Based on acceptable (>90%) TSA control and grass ground cover, triclopyr broadcast or dichlorprop + 2,4-D spot provided the greatest control. With either application method, repeated herbicide applications will be necessary to eliminate TSA because of rapid seedling emergence following control of existing plants.