Nineteenth-century East Africa experienced a first and last, rather than second, efflorescence of slavery. Legal abolition occurred late, between 1897 and 1922. Nevertheless, unlike in many other formerly slave-owning societies, most slave descendants here do not form distinctive, marginalized communities today. Still, they hesitate to acknowledge slave ancestry. This paper investigates the dynamics behind this ambivalent outcome. Comparing two regions in today’s Tanzania, it argues that the role of colonial-era integration into global commodity markets varied between locations, and while it contributed to the obsolescence of slavery, it was neither a sufficient nor a necessary condition for ending slavery and mitigating ex-slave marginality. Rather, ex-slaves’ efforts to acquire unspoiled identities profited from a range of factors, including the chaos of conquest and the First World War, the political and economic repercussions of both these events, and later the depression, on formerly slave-owning elites, and the wide availability of new religious identities. Since a majority of ex-slaves in the region were women, much renegotiation of status occurred within households, relating to markets indirectly.