To evaluate sick leave 12 months after breast cancer surgery, to analyze the effect of adjuvant chemotherapy and to identify predictive factors for sick leave, based on a randomized controlled trial of a non-supervised physical activity intervention (PhysSURG-B). Sick leave days (for patients age 18–67) were collected from the Swedish Social Insurance Agency, and compared between the intervention and control arm in the trial. A random forest model was used to analyze predictive factors for sick leave, including baseline patient and tumor characteristics as well as patient-reported outcomes measuring quality of life, personality traits, pain, anxiety and depression. A total of 203 patients were analyzed, and when comparing intervention to control, there was no differences in sick leave days (117 vs 55 days, odds ratio 1.21; 95% CI 0.75–1.96). Adjuvant chemotherapy significantly increased days of sick leave (323 vs 42 days, odds ratio 17.3; 95% CI 9.15–34.2). Factors predicting prolonged sick leave were adjuvant chemotherapy, young age, previous mental health problems and low quality of life (measured as FACT-B score) at baseline. A non-supervised physical activity intervention had no effect on sick leave 12 months after surgery. Significantly longer sick leave was seen in patients treated by adjuvant chemotherapy. Factors predicting prolonged sick leave can be recognized at baseline and utilized when designing future interventions. • An RCT examining prehabilitation with physical activity in breast cancer surgery. • At 12 months there was no improvement regarding sick leave. • Predictive factors were age, mental health problems and low quality of life. [ABSTRACT FROM AUTHOR]