BACKGROUND: Demographic and infection-related characteristics have been identified as risk factors for long COVID, but research on the influence of health behaviours (e.g., exercise, smoking) immediately preceding the index infection is lacking. The aim of this study was to examine whether specific health behaviours in the month preceding infection with COVID-19 act as upstream risk factors for long COVID as well as well as three specific long COVID symptoms. METHODS: One thousand five hundred eighty-one UK adults from the UCL COVID-19 Social Study and who had previously been infected with COVID-19 were analysed. Health behaviours in the month before infection were weekly exercise frequency, days of fresh air per week, sleep quality, smoking, consuming more than the number of recommended alcoholic drinks per week (> 14), and the number of mental health care behaviours (e.g., online mental health programme). Logistic regressions controlling for covariates (e.g., COVID-19 infection severity, socio-demographics, and pre-existing health conditions) examined the impact of health behaviours on long COVID and three long COVID symptoms (difficulty with mobility, cognition, and self-care). RESULTS: In the month before infection with COVID-19, poor quality sleep increased the odds of long COVID (odds ratio [OR]: 3.53; (95% confidence interval [CI]: 2.01 to 6.21), as did average quality sleep (OR: 2.44; 95% CI: 1.44 to 4.12). Having smoked (OR: 8.39; 95% CI: 1.86 to 37.91) increased and meeting recommended weekly physical activity guidelines (3h hours) (OR: 0.05; 95% CI: 0.01 to 0.39) reduced the likelihood of difficulty with self-care (e.g., washing all over or dressing) amongst those with long COVID. CONCLUSIONS: Results point to the importance of sleep quality for long COVID, potentially helping to explain previously demonstrated links between stress and long COVID. Results also suggest that exercise and smoking may be modifiable risk factors for preventing the development of difficulty with self-care.