The biggest hindrance to clinical application of AAVs is the host immune response that generates neutralising antibodies (NAb) following AAV administration. During vector production, a portion of AAV particles released into the cell culture media is associated with extracellular vesicles (EVs). We hypothesise that these EV-AAVs can provide complete or partial protection from NAbs, allowing for repeated administration and treatment of seropositive patients. However, due to the similar physical properties of EVs and AAVs (i.e size and density) separating EV-AAVs from free AAV particles present in the cell culture media is quite challenging. Five different purification protocols were assessed: differential ultracentrifugation (UC), size exclusion chromatography (SEC), differential gradient centrifugation (DGC), and two combined protocols using two isolation methods, combined protocol 1 and 2 (CP1 and CP2). The presence of EVs and AAV particles was confirmed in all EV-AAV samples using different characterisation methods, however, cryo TEM imaging showed that the level of contamination of EV-AAV samples with free AAV particles was the highest in UC EV-AAV samples (most often used for isolation of EV-AAVs) and the lowest in the CP2 EV-AAV samples. Tested in vitro, CP2 EV-AAV samples showed higher transduction efficiency than AAV or UC EV-AAV when tested in the presence of NAb. Tested in vivo in naive mice (no NAb present in plasma) iv injected with 3*10^11 genome copies per animal, CP2 EV-AAV samples displayed lower transduction efficiency when compared to AAV and UC EV-AAV, but they also displayed lower immunogenicity determined by cell-based NAb assay.