Reports in the 1990s and 2000s showed that kappa opioid receptor (KOR) agonists might be promising for treatment and/or prevention of opioid use disorder (OUD) and cocaine use disorder (CUD). However, the side effects associated with KOR agonists available at the time, such as psychotomimesis, dysphoria and sedation, prevented clinical development. Subsequently, nalfurafine and recently triazole 1.1 and oxa-noribogaine, three centrally acting KOR agonists devoid of such side effects, have been studied in animal models of OUD and CUD. By and large, earlier findings with typical KOR agonists were replicated with nalfurafine and in limited studies with triazole 1.1 and oxa-noribogaine. KOR agonists reduced reinforcing effects of mu opioid receptor (MOR) agonists and decreased tolerance to and dependence on MOR agonists. Oxa-noribogaine suppressed cue-induced reinstatement of morphine and fentanyl seeking. KOR agonists countered itch elicited by MOR agonists and produced additive analgesic effects with MOR agonists, thus allowing use of lower doses of MOR and KOR agonists, resulting in lower degrees of MOR-related side effects (such as respiratory depression) and typical KOR-associated side effects. In addition, KOR agonists attenuated locomotor sensitization and conditioned place preference sensitization following repeated cocaine, reduced acquisition and maintenance of cocaine self-administration and decreased cocaine-induced increase in extracellular dopamine. KOR agonists also suppressed cocaine priming-induced reinstatement of cocaine seeking. Therefore, a combination of a KOR agonist and a MOR agonist or a compound with dual KOR/MOR agonist activities when used as analgesics will deter escalation use of MOR agonists, thus prevent OUD, and KOR agonists may be useful for treatment of cocaine abuse and relapse. Importantly, KOR agonists with no or fewer side effects of typical KOR agonists should be further investigated in animal models of OUD and CUD, particularly those that simulate stress-, cue- and drug priming-induced relapse for potential clinical development.