Background: Levels of self-reported health do not always correlate with levels of physical disability in stroke survivors. We aimed to explore what underlies the difference between subjective self-reported health and objectively measured disability among stroke survivors. Methods: Face to face semi-structured interviews were conducted with stroke survivors recruited from a stroke clinic or rehabilitation ward in the UK. Fifteen stroke survivors purposively sampled from the clinic who had discordant self-rated health and levels of disability i.e. reported health as ‘excellent’ or ‘good’ despite significant physical disability (eight), or as ‘fair’ or ‘poor’ despite minimal disability (seven) were compared to each other, and to a control group of 13 stroke survivors with concordant self-rated health and disability levels. Interviews were conducted 4 to 6 months after stroke and data analysed using the constant comparative method informed by Albrecht and Devlieger’s concept of ‘disability paradox’. Results: Individuals with ‘excellent’ or ‘good’ self-rated health reported a sense of self-reliance and control over their bodies, focussed on their physical rehabilitation and lifestyle changes and reported few bodily and post-stroke symptoms regardless of level of disability. They also frequently described a positive affect and optimism towards recovery. Some, especially those with ‘good’ self-rated health and significant disability also found meaning from their stroke, reporting a spiritual outlook including practicing daily gratitude and acceptance of limitations. Individuals with minimal disability reporting ‘fair’ or ‘poor’ self-rated health on the other hand frequently referred to their post-stroke physical symptoms and comorbidities and indicated anxiety about future recovery. These differences in psychological outlook clustered with differences in perception of relational and social context including support offered by family and healthcare professionals. Conclusions: The disability paradox may be illuminated by patterns of individual attributes and relational dynamics observed among stroke survivors. Harnessing these wider understandings can inform new models of post-stroke care for evaluation.