Introduction During the past decade, substantial evidence has documented that opioid receptor heterodimers form in cell lines expressing one or more opioid receptors. More recent studies have begun to investigate whether heterodimer formation also occurs in vivo. Objectives We have used opioid receptor knockout mice to determine whether the in vivo intrathecal (i.t.) pharmacological potency of delta, kappa, and bivalent kappa/delta ligands is altered in the absence of the KOR-1 and/or DOR-1 genes. Results We observe that both NorBNI (a kappa antagonist) and KDN-21 (a kappa/delta, bivalent antagonist) specifically inhibit DPDPE but not deltorphin II i.t potency in wildtype mice but that, following mutation of KOR-1, the ability of either compound to reduce DPDPE potency is lost. In contrast, knockout of KOR-1 unexpectedly slightly reduces the potency of deltorphin II (delta2) but not DPDPE (delta1). Finally, two compounds with kappa, agonist activity, 6′-GNTI (a putative kappa/delta heterodimer selective agonist) and KDAN-18 (kappa agonist/delta antagonist bivalent ligand) show reduced potency in DOR-1 KO mice. Conclusions These results show, genetically, that bivalent ligands with kappa agonist activity require delta receptors for maximal potency in vivo, which is consistent with the presence of opioid heterodimer/oligomer complexes in vivo, and also highlight the complexity of delta drug action even when complementary pharmacologic and genetic approaches are used.