This scoping review intervenes in a fast-growing yet fragmented field. Sex dolls and sex robots are cast as artefacts that promise the personalization of intimacy (Illouz 2007; 2012), while scholarship ranges across ethical–normative debates, STS/posthumanist perspectives, the sociology of emotions, and sexual-script theory under heterogeneous labels that hinder comparison. We ask how sociology has conceptualized and labelled these artefacts (2000–2025), which theoretical frames and lexicons predominate and how they have evolved, and which methodological approaches are privileged alongside the main thematic, geographical, and methodological gaps. Following Arksey & O’Malley (2005) and PRISMA-ScR (Tricco et al., 2018), we search Scopus and Web of Science using strings that combine “sex robot,” “love doll,” “real doll,” “sex doll,” and “companion robot.” We include peer-reviewed contributions adopting a social perspective that treat dolls/robots as a primary object or a relevant case, and exclude studies that are purely clinical/medical, legal, engineering, or psychological without sociocultural analysis. Data are organized in an extraction matrix (Excel) and analysed with Biblioshiny to provide an initial quantitative mapping and a qualitative, directed content analysis that identifies and defines the field’s core conceptual frames. The study starts in 2000 to capture shifts that precede and accompany two symbolic milestones: David Levy’s Love and Sex with Robots (2007) and the public unveiling of Roxxxy at the AVN Adult Entertainment Expo (Las Vegas, 9 January 2010). A prior synthesis (Döring & Pöschl, 2018) offers interdisciplinary context but is not a sociological scoping review and does not systematically map theoretical frames—gaps this study addresses. Preliminary patterns indicate a predominance of ethical/speculative discourse over empirical research, Western/Anglophone concentration, non-standardized terminology, limited attention to everyday practices (care, partner negotiation) and to disability/LGBTQ+ perspectives, and few mixed-methods designs. Expected outputs are a reproducible corpus, a typology of sociological frames, and an indicator matrix to operationalize core constructs, improving comparability and cumulative knowledge.