Spinocerebellar ataxia type 3 (SCA3) is a cureless neurodegenerative disease recognized as the most prevalent form of dominantly inherited ataxia worldwide. The main hallmark of SCA3 is the expansion of a polyglutamine tract located in the C-terminal of Ataxin-3 (or ATXN3) protein, that triggers the mis-localization and toxic aggregation of ATXN3 in neuronal cells. The propensity of wild type and polyglutamine-expanded ATXN3 proteins to aggregate has been extensively studied over the last decades. In vitro studies with mass spectrometry techniques revealed a time-dependent aggregation of polyglutamine-expanded ATXN3 that occurs in several steps, leading to fibrils formation, a high status of aggregation. For in vivo experiments though, the techniques commonly used to demonstrate aggregation of polyglutamine proteins, such as filter trap assays, SDS-PAGE and SDS-AGE, are unable to unequivocally show all the stages of aggregation of wild type and polyglutamine-expanded ATXN3 proteins. Here we describe a systematic and detailed analysis of different known techniques to detect the various forms of both wild type and pathologic ATXN3 aggregates, and we discuss the power and limitation of each strategy.