Testosterone concentrations are lower in older compared to younger men, representing both a biomarker of and a contributor to poorer health. Sociodemographic, lifestyle, behavioural and medical factors are associated with differences in testosterone concentrations. Recent analyses from UK Biobank and the Androgens In Men Study collaboration indicate that healthy men aged 40-69 years have stable testosterone concentrations over four years of follow-up, but beyond the age of 70 years, declining testosterone concentrations are accompanied by increasing luteinising hormone, indicative of Leydig cell impairment. In UK Biobank men, lower testosterone concentrations were independently associated with mortality, and with incident dementia, but not with incident cardiovascular events. The association of lower testosterone with higher mortality risk in UK Biobank and in the Health In Men Study was non-linear, with no survival advantage for men with higher testosterone concentrations. These results suggest that an optimal testosterone concentration, or range of concentrations, exists, which predicts male health and longevity. In large randomised controlled trials (T Trials, T4DM and TRAVERSE) testosterone treatment improved sexual function, anaemia and bone density, and reduced the risk of type 2 diabetes, with no increase in cardiovascular adverse events, but an increase in fracture risk. Further research is needed to clarify the extent to which non-pharmacological interventions might help men transitioning from middle to older age attain both optimal circulating testosterone concentrations and better health.