Beta-arrestins are key negative regulators and scaffolds of G protein-coupled receptor (GPCR) signalling. Beta-arrestin1 and beta-arrestin2 preferentially bind to the phosphorylated GPCRs in response to agonist stimulation, resulting in receptor internalization and desensitization. The critical roles of GPCR kinases (GRKs)-catalyzed receptor phosphorylation and interaction of beta-arrestins with the phosphorylated receptor in receptor internalization are well established. However, emerging evidence suggests that an agonist-stimulated internalization mechanism that is independent of receptor phosphorylation may also be employed in some cases, although the molecular mechanism for the phosphorylation-independent GPCR internalization is not clear. The current study investigated the role of receptor phosphorylation and the involvement of different beta-arrestin subtypes in agonist-induced delta-opioid receptor (DOR) internalization in HEK293 cells. Results from flow cytometry, fluorescence microscopy, and surface biotin labelling experiments showed that elimination of agonist-induced DOR phosphorylation by mutation GRK binding or phosphorylation sites only partially blocked agonist-induced receptor internalization, indicating the presence of an agonist-induced, GRK-independent mechanism for DOR internalization. Fluorescence and co-immunoprecipitation studies indicated that both the wild-type DOR and the phosphorylation-deficient mutant receptor could bind and recruit beta-arrestin1 and beta-arrestin2 to the plasma membrane in an agonist-stimulated manner. Furthermore, internalization of both the wild-type and phosphorylation-deficient receptors was increased by overexpression of either type of beta-arrestins and blocked by dominant-negative mutants of beta-arrestin-mediated internalization, demonstrating that both phosphorylation-dependent and -independent internalization require beta-arrestin. Moreover, double-stranded RNA-mediated interference experiments showed that either beta-arrestin1 or beta-arrestin2 subtype-specific RNAi only partially inhibited agonist-induced internalization of the wild-type DOR. However, agonist-induced internalization of the phosphorylation-deficient DOR was not affected by beta-arrestin1-specific RNAi but was blocked by RNAi against beta-arrestin2 subtype. These data indicate that endogenous beta-arrestin1 functions exclusively in the phosphorylation-dependent receptor internalization, whereas endogenous beta-arrestin2, but not beta-arrestin1, is required for the phosphorylation-independent receptor internalization. These results thus provide the first evidence of different requirement for beta-arrestin isoforms in the agonist induced phosphorylation-dependent and -independent GPCR internalization.