Introduction The nature of the relationship between country-level energy consumption and life expectancy has not been examined systematically.

Methods Ecological study of carbon dioxide (CO₂) emissions per capita and life expectancy at birth using the maximum available sample of countries.

Results Among 155 countries CO₂ emissions per capita were in the range 0–30 metric tons per year. Life expectancy at birth rose sharply between 0 and 5 t of emissions from 42 (Afghanistan) to 78 (Costa Rica, Chile). Spearman's rank correlation was 0.78, p<0.0001. At higher levels of CO₂ emissions there was no association with life expectancy. Among 30 countries with high life expectancy (>75 years) and relatively low CO₂pc emissions (<10 t) with a population>0.5 M there was a modest association overall between CO₂ emissions and life expectancy (rank correlation 0.51, p=0.004). Within this group, life expectancy ranged between 75.5 (Argentina, annual emissions 4.6 t) and 82.5 years (Switzerland, 5.1 t). Using the 2008 World Bank income classification, per capita emissions were associated with life expectancy among low and low middle income countries, but not in high middle or high income countries.

Conclusions Life expectancy, a surrogate indicator of population health, is unrelated to energy consumption above a low level. It appears, looking to the future, that the environmental sustainability and global health development agendas are compatible in practice.