Serum testosterone concentrations decline with age, while serum sex hormone-binding globulin (SHBG) concentrations increase.To analyse associations of baseline serum testosterone and SHBG concentrations, and calculated free testosterone (cFT) values, with all-cause and cause-specific mortality in men.The UK Biobank prospective cohort study of community-dwelling men 40-69 years-old, followed for 11 years.All-cause, atherosclerotic cardiovascular disease (CVD) and cancer-related mortality. Cox proportional hazards regression was performed, adjusting for age, waist circumference, medical conditions and other covariates. Models for testosterone included SHBG, and vice versa.In complete case analysis of 149,436 men with 10,053 deaths (1,925 CVD and 4,927 cancer-related), men with lower testosterone had higher mortality from any cause (lowest vs highest quintile, Q1 vs Q5, fully-adjusted hazard ratio [HR]=1.14, 95% confidence interval [CI]=1.06-1.22, overall trend P<0.001), and cancer (HR=1.20, CI=1.09-1.33, P<0.001), with no association for CVD deaths. Similar results were seen for cFT. Men with lower SHBG had lower mortality from any cause (Q1 vs Q5, HR=0.68, CI=0.63-0.73, P<0.001), CVD (HR=0.70, CI=0.59-0.83, P<0.001), and cancer (HR=0.80, CI=0.72-0.89, P<0.001). A multiply-imputed dataset (N=208,425, 15,914 deaths, 3,128 CVD and 7,468 cancer-related) and analysis excluding deaths within first two years (9,261, 1,734 and 4,534 events) yielded similar results.Lower serum testosterone is independently associated with higher all-cause and cancer-related, but not CVD-related, mortality in middle-aged to older men. Lower SHBG is independently associated with lower all-cause, CVD-related and cancer-related mortality. Confirmation and determination of causality requires mechanistic studies and prospective trials.