Population-based data regarding the prognostic implication of gamma-glutamyl transferase (GGT) have been inconsistent. We examined the association of GGT with all-cause and disease-specific mortality. Using the Korean nationwide database, we included 9,687,066 subjects without viral hepatitis or cirrhosis who underwent a health examination in 2009. Subjects were classified into three groups by sex-specific tertile of serum GGT levels. The underlying causes of death were classified by 10th Revision of the International Classification of Diseases codes. During the median follow-up period of 8.3 years, 460,699 deaths were identified. All-cause mortality increased as serum GGT levels became higher (hazard ratio [HR], 95% confidence interval [CI] 1.05, 1.04–1.05 in the middle tertile, and 1.33, 1.32–1.34 in the high tertile) compared to the low tertile of serum GGT levels. Similar trends were observed for cardiovascular disease (CVD) (HR, 95% CI 1.07, 1.05–1.09 in the middle tertile, 1.29, 1.26–1.31 in the high tertile), cancer (HR, 95% CI 1.08, 1.07–1.10 in the middle tertile, 1.38, 1.36–1.39 in the high tertile), respiratory disease (HR, 95% CI 1.10, 1.08–1.13 in the middle tertile, 1.39, 1.35–1.43 in the high tertile), and liver disease mortality (HR, 95% CI 1.74, 1.66–1.83 in the middle tertile, 6.73, 6.46–7.01 in the high tertile). Regardless of smoking, alcohol consumption and history of previous CVD and cancer, a higher serum GGT levels were associated with a higher risk of mortality. Serum GGT levels may be useful for risk assessment of all-cause and disease-specific mortality in general population.