When the bacillus Calmette-Guérin (BCG) vaccine was introduced in the 1920s, it was suggested that BCG occasionally had nonspecific beneficial effects on mortality beyond the specific protection against tuberculosis. Considering that BCG has since then become the most used vaccine in the world, surprisingly few studies have been undertaken into the effect of BCG on general mortality and morbidity. Recent studies suggest that BCG has beneficial nontargeted effects on general infant morbidity and mortality in low-income countries, often with the most pronounced effect among girls. These observational findings are supported by early trials in which children were randomized or alternated to BCG vaccination. Furthermore, a BCG scar and a positive tuberculin reaction are related to better survival among BCG-vaccinated children in low-income countries, especially for girls. The findings are not explained by frailty bias, in other words, that healthy children are more likely to receive BCG vaccination. A nonspecific, gender-differential effect of BCG on general infant mortality may have large implications for tuberculosis vaccine research and routine vaccination policy.