Physical activity (PA) is one of the most powerful treatment options for persons with prediabetes or diabetes. However, some elevation in risk occurs with increased PA, at least initially, and certain precautions need to be made to lower these risks, particularly if these persons are unaccustomed to exercise. We conducted a standardized search of all adverse events associated with increased PA in persons with prediabetes or diabetes (type 1 or type 2) and provided evidence-based guidelines on PA screening in these apparently high-risk individuals. A systematic literature review was performed of all studies reporting on adverse events in persons with prediabetes or diabetes. Studies included were from all designs (retrospective and prospective including randomized controlled trials) and were assessed according to evaluation criteria adapted by a consensus panel. A total of 47 studies, involving >8000 individuals, were deemed eligible. A number of these studies identified a range of mild to severe acute risks with exercise (musculoskeletal injury, hypoglycemia, foot ulceration, proliferative retinopathy, hypotension, sudden death) but the overall prevalence was low. Based on several randomized controlled trials and prospective studies in which prescribed exercise was performed at a wide range of intensities, it appears that increased PA is a relatively safe procedure with no evidence of a loss of life. Based on our assessment of the available literature, we provide a new PA risk algorithm for persons with prediabetes and diabetes and comment on the role of the patient, the qualified exercise professional, and the patient's physician in the risk screening process.