Child Mortality Through History
In 1800, roughly 43% of children worldwide died before their fifth birthday. In 1950, the figure was 22%. In 2024, it is 3.7%. The absolute number of under-5 deaths has fallen from over 18 million a year in 1950 to 4.9 million today — a decline matched by very few measurable improvements in human welfare.
Key insights
Sanitation, not medicine, did most of the work
Most of the historical decline in child mortality came before antibiotics and most modern vaccines. Clean water, sewerage, soap, refrigeration, better nutrition and maternal education drove the 19th and early-20th century improvements. Pathogens that thrive in dirty water — typhoid, cholera, dysentery — were the dominant killers and the cheapest to prevent.
Vaccines and antibiotics accelerated post-1950 gains
Mass vaccination (measles, DTP, polio), oral rehydration therapy, antibiotic treatment of pneumonia, insecticide-treated bed nets and PMTCT for HIV each cut major causes of child death substantially. The expanded programme on immunization (EPI), launched 1974, has prevented an estimated 154 million deaths cumulatively.
The gap by country is still large
Country-level under-5 mortality ranges from under 0.3% (Singapore, Japan, Finland) to over 11% (Nigeria, Somalia, Niger, Chad). The 30-fold national gap dwarfs anything seen in other social indicators. Most of the remaining 4.9 million annual deaths are concentrated in 30 countries; pneumonia, diarrhoea, malaria and neonatal complications still cause two thirds of them.
Global under-5 mortality 1800–2024
% of children dying before age 5
Key Finding: From 43% in 1800 to 3.7% today — a more than 90% decline.
Under-5 mortality — selected countries 2024
Deaths before age 5 per 1,000 live births
Key Finding: The 30-fold gap between best and worst persists despite global progress.
Methodology & caveats
The under-5 mortality rate
The U5MR is the probability that a child born in a specified year will die before age 5, expressed per 1,000 live births. It is a probabilistic measure derived from age-specific death rates, not a simple count. Civil registration (birth/death certificates) provides the cleanest data; in many low-income countries it comes from demographic and health surveys.
Historical reconstructions
Pre-1950 estimates rely on parish records, family reconstitution studies, and historical demography techniques (Sweden has the longest continuous series, back to 1750). Pre-1800 estimates are less reliable but converge on roughly 40–50% under-5 mortality across the pre-industrial world. The IGME assembles a harmonized historical series.
Neonatal vs post-neonatal
Today, about half of under-5 deaths happen in the first 28 days of life (neonatal mortality) — driven by prematurity, birth complications and neonatal infections. The remaining deaths cluster around pneumonia, diarrhoea and malaria. As programmes have reduced post-neonatal mortality, the neonatal share has risen and become the harder problem to address.