Experiments were conducted to evaluate the toxicity, feeding preference, repellency, and field efficacy associated with the organic insecticides azadirachtin, pyrethrins, and spinosad against two stink bug species, Acrosternum hilare (Say) and Euschistus servus (Say) (Hemiptera: Pentatomidae). Laboratory toxicity bioassays were conducted using treated green bean pods. The conventional pyrethroid lamda-cyhalothrin was included for comparison. A. hilare adults and nymphs were most susceptible to lamda-cyhalothrin and to tank mixes of pyrethrins + spinosad. E. servus adults were susceptible to lamda-cyhalothrin, spinosad, and all tank mixes, whereas E. servus nymphs were susceptible to lamda-cyhalothrin only. Feeding preference tests were conducted using insecticide-treated tomatoes and counting the number of feeding stylet sheaths on fruit after 24 h. All tomatoes treated with either azadirachtin, pyrethrins, or tank mixes resulted in fewer numbers of stylet sheaths than the untreated control, whereas treatment with spinosad alone did not. In filter paper repellency tests, both E. serous and A. hilare were repelled by pyrethrins and exhibited no response to azadirachtin. E. servus was attracted to spinosad in comparison with a water-treated control; however, A. hilare displayed no response. In field efficacy trials, each of the organic insecticides reduced the number of stink bugs in soybean, Glycine max (L.) Merr., for up to 2 d after treatment; however, none of the insecticides reduced stink bug damage to fruit in tomatoes even after multiple applications. Implications for organic growers and integrated pest management programs are discussed.