Background: Exercise referral schemes (ERS) are prescribed programs to tackle physical inactivity and associated noncommunicable disease. Inconsistencies in reporting, recording, and delivering ERS make it challenging to identify what works, why, and for whom. Methods: Preferred Reporting Items for Systematic Reviews and Meta-Analyses guided this narrative review of reviews. Electronic databases were searched for systematic reviews of ERS. Inclusion criteria and quality assessed through A Measurement Tool to Assess Systematic Reviews (AMSTAR). Data on uptake, attendance, and adherence were extracted. Results: Eleven reviews met inclusion criteria. AMSTAR quality was medium. Uptake ranged between 35% and 81%. Groups more likely to take up ERS included (1) females and (2) older adults. Attendance ranged from 12% to 49%. Men were more likely to attend ERS. Effect of medical diagnosis upon uptake and attendance was inconsistent. Exercises prescribed were unreported; therefore, adherence to exercise prescriptions was unreported. The influence of theoretically informed approaches on uptake, attendance, and adherence was generally lacking; however, self-determination, peer support, and supervision were reported as influencing attendance. Conclusions: There was insufficient reporting across studies about uptake, attendance, and adherence. Complex interventions such as ERS require consistent definitions, recording, and reporting of these key facets, but this is not evident from the existing literature.