Schizophrenia is a debilitating chronic mental illness that affects people’s ability to perform daily activities. Schizophrenia is treated with antipsychotics and psychotherapy which only target the symptoms, however there is little known about its molecular background. There are several biological theories of schizophrenia, but the most influential is the dopamine hypothesis, which states that there is an imbalance in dopamine transmission. Other theories suggest neurodevelopmental changes, such as brain maturation and altered neurotransmission, as key factors in development of schizophrenia. Along with genetic and environmental factors, which contribute to development of schizophrenia, protein aggregation was introduced as a possible non-genetic cause. Protein aggregation occurs when proteins misfold due to errors in folding or trafficking machinery. One newly discovered protein to aggregate in brains of schizophrenia patients is EH-domain containing 3 (EHD3). EHD3 is believed to be one of the main regulators of endocytic recycling involved, among other functions, in endosome trafficking and promotion of dopaminergic transmission. Previous research showed that full length EHD3 consistently aggregates in SH-SY5Y cells, while a truncated form EHD3 1-399, with its linker region deleted, does not. In contrast, EHD3 1-434, which contains both the linker region and an EH domain, aggregates. Linker region was therefore suspected to be a critical region for EHD3 aggregation. In this thesis, we aimed to determine whether the aggregation occurs within 400- 434 amino acids of EHD3 by fluorescent microscopy and ultracentrifugation assay. Ultracentrifugation data surprisingly revealed that EHD3 1-434 does not appear in the aggregome of HEK293 cells, as hypothesized, but it still formed aggregate-like structures in both SH-SY5Y and HEK293 cells. Since EHD3 1-434 probably does not aggregate, we suspected involvement of the EH domain in EHD3 aggregation. Novel fragments EHD3 1-404 and EHD3 400- 535 were therefore cloned and expressed in both cell lines. However, they did not show any signs of aggregation. Co-aggregation of full length EHD3 with non-aggregating EHD3 1-399 was also investigated. Fluorescent microscopy results showed that full length EHD3 induces EHD3 1-399 to aggregation in neuroblastoma cells. Future research is needed to determine the critical region of EHD3 aggregation by generating novel constructs that include small parts of linker region or the EH domain.