Abstract. The full-text article is available at http://www.effective-states.org/working-paper-124/In recent decades, reforms have been introduced in developing countries to promote economic transformation, democracy and the rule of law. However, structural factors have often undermined their implementation. This is a key insight of the political settlement analysis that has proliferated in scholarly research. It is unpacking of the sorts of intra-elite relations that are instrumental in choosing policies and their modes of implementation is a major achievement. However, with its focus on hard force and economic rents, it is less clear regarding the role of elections and popular legitimacy, which have become more important recently. Inspired by an adapted political settlement analysis, and by drawing on the strategic-relational approach, this paper aims to explain contemporary forms of power and legitimacy in greater detail. Using Tanzania – which has had the same party in power since independence – as a critical case study, we demonstrate that, in the context of democratisation, the country’s political elites are increasingly attempting to earn popular legitimacy. In Tanzania, earlier attempts to earn popular legitimacy through the expansion of social services to the rural majority were radicalised when a new president came to power in 2015. During the historically competitive elections, he campaigned on a platform of reversing years of domination by business and political elites. He later crafted a series of nationalist narratives and attacks on private investors, not least foreign ones, to bolster his legitimacy in the eyes of the wider population. This implies a more prominent role for populations in developing countries than is often acknowledged. We also suggest that, in the context of democratisation, analyses of legitimacy should include two more dimensions: first, a political elite’s relationship with its political opponents, who in Tanzania have been systematically delegitimised; and secondly, international recognition, which since the 1980s has required the holding of regular elections and is important for resource mobilisation. We, therefore, argue that legitimacy should be analysed as a source of power in its own right, in line with force and rents; it is the combination of these different sources of power that matters.