PurposeThe purpose of this study is to examine how accountability is constructed for blockchain systems. With the aim of increasing knowledge on accountability across three different types of blockchains (public, private and consortium), the researchers ask: how do blockchain systems construct accountability?Design/methodology/approachThis study draws on theorising in the accountability literature to study how blockchains relate to our construction and understanding of accountability. A qualitative field study of the Australian blockchain technology landscape is conducted, with insights garnered from 18 blockchain experts.FindingsFindings reveal that different types of blockchains employ different forms and mechanisms of accountability and in novel ways previously less acknowledged in the literature. Importantly, this study finds that accountability does not require a principal–agent relation and can still manifest in less pure applications of blockchain technology across a wide range of stakeholders, contrary to that espoused in earlier exhortations of blockchain use in interdisciplinary literature. This study also finds that similar subtypes of accountability operate very differently across public, private and consortium blockchains and there exists an inverse relation between trust and consensus building through transparency as blockchains progress from public to private types. Overall, this study offers novel explanations for the relevance of greater accountability in blockchains, especially when the assumptions of public blockchains are softened and applied as private and consortium blockchains.Originality/valueThis study contributes to the accountability literature by addressing how different blockchain systems reshape the understanding of traditional accounting and accountability practices. This study questions the very need for a principal–agent relation to facilitate accountability and offers an additional perspective to how trust and transparency operate as key mechanisms of accountability.