Background Physical deconditioning in combination with societal and emotional factors has been hypothesized to compromise complete recovery from low back pain (LBP). However, there is a lack of longitudinal studies designed to specifically investigate physical activity as an independent prognostic factor. We conducted a prognostic study to investigate whether levels of leisure time physical activity are independently associated with clinical outcomes in people seeking care for chronic and persistent LBP. Methods A total of 815 consecutive patients presenting with LBP to an outpatient spine centre in secondary care were recruited. Separate multivariate linear regression analyses were performed to investigate whether levels of leisure time physical activity (i.e., sedentary, light and moderate-to-vigorous leisure time physical activity levels) predict pain and disability at 12-month follow-up, after adjusting for age, pain, episode duration, disability, neurological symptoms, depression and fear of movement. Results Final models showed evidence of an association between baseline physical activity and 12-month outcomes (p < 0.001). In both models, the moderate-to-vigorous physical activity group reported less pain and disability compared with the sedentary group. Conclusions Our findings suggest that physical activity levels may have a role in the prognosis of LBP. Specific domains of physical activity warrant further investigation to better understand this association.