Objectives: To determine the personal and social factors perceived to influence physical activity levels in stroke survivors. Data sources: Four electronic databases (MEDLINE, CINAHL, PubMed and Embase) were searched from inception to November 2020, including reference and citation list searches. Study selection: The initial search yielded 1499 papers, with 14 included in the review. Included articles were peer-reviewed, qualitative studies, reporting on the perceived factors influencing physical activity levels of independently mobile community-dwelling adults, greater than 3 months post stroke. Data extraction: Data extracted included location, study aim, design, participant and recruitment information and how data were collected and analysed. Data synthesis: Thematic analysis was undertaken to identify meanings and patterns, generate codes and develop themes. Five main themes were identified: (i) Social networks are important influencers of physical activity; (ii) Participation in meaningful activities rather than ‘exercise’ is important; (iii) Self-efficacy promotes physical activity and physical activity enhances self-efficacy; (iv) Pre-stroke identity related to physical activity influences post-stroke physical activity; and (v) Formal programmes are important for those with low self-efficacy or a sedentary pre-stroke identity. Conclusions: Physical activity levels in stroke survivors are influenced by social activities and support, pre-stroke identity, self-efficacy levels and completion of activities that are meaningful to stroke survivors.