The paper reports an evidence synthesis of how loneliness is conceptualised in qualitative studies in adults. Using PRISMA guidelines, our review evaluated exposure to or experiences of loneliness by adults (aged 16+) in any setting as outcomes, processes, or both. Our initial review included any qualitative or mixed-methods study, published or unpublished, in English, from 1945 to 2018, if it employed an identified theory or concept for understanding loneliness. The review was updated to include publications up to November 2020. We used a PEEST (Participants, Exposure, Evaluation, Study Design, Theory) inclusion criteria. Data extraction and quality assessment (CASP) were completed and cross-checked by a second reviewer. The Evidence of Reviews of Qualitative Research (CERQual) was used to evaluate confidence in the findings. We undertook a thematic synthesis using inductive methods for peer-reviewed papers. The evidence identified three types of distinct but overlapping conceptualisations of loneliness: social, emotional, and existential. We have high confidence in the evidence conceptualising social loneliness and moderate confidence in the evidence on emotional and existential loneliness. Our findings provide a more nuanced understanding of these diverse conceptualisations to inform more effective decision-making and intervention development to address the negative wellbeing impacts of loneliness.