The included studies were of moderate to high quality, though some lacked detailed reporting on blinding or allocation concealment. Analysis suggests that the timing of intervention initiation, along with exercise type and frequency, may contribute to observed variations in outcomes. Evidence quality assessments did not reveal significant issues with indirectness or imprecision, and no significant publication bias was detected. Given the low heterogeneity and absence of significant downgrade factors, intermediate evidence quality was assigned for upper limb function and shoulder abduction, with high quality for shoulder flexion. Physical exercise is notably effective in enhancing both upper limb function and shoulder mobility in breast cancer patients, with the timing and frequency of exercise interventions influencing these improvements. This provides valuable evidence for clinical rehabilitation strategies.