Ageing is accompanied by a reduction in circulating testosterone (T), and progressive accumulation of medical morbidities. There is ongoing debate as to whether low T contributes to ill-health, particularly to increased risk of cardiovascular disease, as opposed to being a biomarker for its presence. Despite this uncertainty, prescriptions for T are rising on a background of concern over potential adverse effects. Observational studies show lower risk of cardiovascular events in older men with higher T concentrations. In longitudinal analyses from the Western Australian Health In Men Study (HIMS) we have shown that optimal circulating T predicts survival in older men, and that higher T concentrations are independently associated with reduced incidence of stroke. Furthermore, in HIMS men with higher concentrations of the more potent androgen dihydrotestosterone (DHT) experienced lower mortality from ischaemic heart disease. Concern has been raised following the Testosterone in Older Men with Mobility Limitations (TOM) trial, which was terminated due to excess cardiovascular adverse events reported in the T treatment arm. However, no such signal was seen in a comparable study of T in intermediate-frail and frail older men. Of note, these and other randomised controlled trials (RCTs) of T supplementation have been underpowered for the outcome of cardiovascular events. Recent meta-analyses generally have not shown an excess of cardiovascular adverse events to be associated with T therapy. Retrospective studies of prescription databases have produced controversial and conflicting results. Thus additional RCTs are required to clarify the role of T supplementation to modulate cardiovascular risk in older men in the absence of pituitary or testicular disease. T replacement therapy should be considered in androgen deficient men, with evaluation of potential benefits and risks.