The First World War transformed a European crisis into a conflict fought across continents and oceans. Mass armies depended on factories, railways, shipping and financial systems, while civilians faced occupation, shortages and displacement. These statistics examine the scale of mobilisation, the human losses, the overlapping campaigns and the resources required to sustain them. They also show why historical numbers need context: a casualty is not necessarily a death, a mobilisation total is not an army’s strength on one day, and published loss estimates can differ substantially. Each chart identifies its coverage, with a reference table bringing the main belligerents together.
Explore the charts by hovering, tapping or using their selectors. Historical estimates and territorial definitions vary; each chart explains its coverage.
Statistics
- Military deaths relative to pre-war population
- How many people were mobilised?
- A war fought on overlapping fronts
- British merchant shipping lost each month
- Artillery production during the war
- Casualties in major Western Front battles
- Where British Empire army troops came from
- National debt before and after the war
- Military deaths and civilian death estimates
- When countries entered and left the war
- Major belligerents: reference table
1. Military deaths relative to pre-war population
Hover, tap or choose an item for detailed values.
Military deaths / pre-war population: ■ under 0.5% ■ 0.5–under 1% ■ 1–under 2% ■ 2–under 4% ■ 4%+ Grey: no figure. Approximate historical boundaries.
Absolute death totals reveal the size of a loss; population-based rates reveal a different part of its impact. This map divides estimated military deaths by the pre-war population used in a published comparative study. Small countries can therefore appear prominently even when their total losses were lower than those of the largest powers. The denominator includes the whole population, not just men of military age or people who served. The map uses a reconstruction of 1914 boundaries because modern borders would split several wartime empires into countries that did not then exist in their present form. Serbia and Montenegro share a combined estimate in the source and are coloured accordingly. Colonial losses are not silently assigned the metropolitan country’s rate. Grey areas have no matching value in this comparison. Select a country for the numerator, denominator and resulting percentage, and switch between the European view and the wider world.
Sources: Broadberry & Harrison, The Economics of World War I, Table 9; Historical basemaps: approximate 1914 borders (GPL-3.0). Historical country categories from Broadberry & Harrison, Table 9, drawing on Urlanis. Military deaths divided by pre-war population; not total war deaths. UK includes Ireland. Serbia and Montenegro share one combined rate. Colonial troops are separate in the source and are not assigned metropolitan population denominators. Boundaries are approximate, not a territorial settlement map.
2. How many people were mobilised?
Mobilisation brought people into military service on a scale that cannot be understood from the size of a single battle. These bars group the published national estimates into the Allied and Associated Powers and the Central Powers. Each coloured segment represents one country or empire, while the full bar shows the sum of the included entries. The British Empire appears as one entry here, so its dominions and India are not added again. France and other imperial powers retain the geographical coverage of the underlying historical table. These are cumulative mobilisation estimates, not the number simultaneously serving at the front. Some personnel served far from combat, and the source’s national recording conventions were not identical. The overall comparison is useful for scale, but it does not measure training, equipment, deployment speed or combat effectiveness. Use the country selector to inspect small contributions without having to hit a narrow segment.
Sources: Britannica: mobilisation and losses. Britannica reproduces U.S. War Department estimates from February 1924. Allied and Associated total 42,188,810; Central Powers 22,850,000. British Empire appears once. Country selector isolates a contribution; reset to All countries for alliance totals. Mobilisation totals are estimates, not simultaneous frontline strength.
3. A war fought on overlapping fronts
Hover, tap or choose an item for detailed values.
The familiar image of the Western Front captures only one part of the war. This timeline places selected campaigns on a shared calendar, making their overlap immediately visible. Fighting in Africa and the Middle East often continued while European offensives dominated public attention, placing competing demands on soldiers, ships, medical services and supplies. Each bar follows the stated campaign boundaries rather than the period during which every participating soldier was present. The Gallipoli bar covers the land campaign, for example, while the Russian Eastern Front extends to the peace settlement after an earlier interruption in fighting. Campaigns also had quieter intervals that a continuous bar cannot describe. The East African endpoint falls after the European armistice, illustrating why 11 November did not instantly end hostilities everywhere. Hover or select a campaign to see its dates and the event used to define each endpoint.
Sources: National Army Museum timeline. Selected campaigns with conventional opening and closing dates. Bars indicate elapsed time, not intensity or uninterrupted fighting. The Russian timeline ends with the March 1918 peace and includes the preceding armistice. East African hostilities continued beyond 11 November. This is not an exhaustive campaign inventory.
4. British merchant shipping lost each month
Merchant shipping connected Britain with food, raw materials and the wider imperial war effort. The monthly series shows the gross tonnage of British merchant vessels lost through enemy action, with a particularly severe peak in April 1917. Gross tonnage measures internal carrying volume rather than the weight of cargo aboard or the number of people killed. The source includes enemy action generally, so these figures should not be described as losses caused exclusively by submarines. Fishing vessels are recorded separately in the original table and are excluded here. The dates of unrestricted submarine warfare and the expansion of convoy protection help place the changes in context, but a falling line alone cannot isolate the effect of any single policy. The final observation covers only the first eleven days of November 1918. That shortened period is marked explicitly so its small value is not mistaken for a full month’s result.
Sources: Archibald Hurd, The Merchant Navy, Volume III, Appendix C. Hurd, The Merchant Navy, Volume III, Appendix C. All 52 months from August 1914 through November 1918 are present. British merchant vessels lost through enemy action, not solely submarines. Excludes fishing vessels, neutral and other Allied shipping, and marine accidents. Gross registered tonnage measures enclosed volume, not cargo weight. Final month covers 1–11 November only.
5. Artillery production during the war
An army’s ability to keep fighting depended on replacing weapons as well as recruiting soldiers. This comparison follows annual artillery production in Britain and Austria-Hungary, two belligerents with a complete five-year series in the selected compilation. It measures artillery pieces, not shells, expenditure or the number of guns already available when war began. Keeping one physical unit avoids the misleading practice of combining rifles, aircraft and ammunition into a single production count. The figures show contrasting production paths, but they do not reveal every difference in industrial capability. A light field gun and a heavy howitzer count as one piece each, although their manufacturing requirements and battlefield roles differ. Imports, repairs, captured equipment and changes in quality are also outside this simple count. The 1914 and 1918 observations belong to years in which the war occupied only part of the calendar, another reason to read annual comparisons with care.
Sources: Ellis & Cox production figures, reproduced by WorldWars.net. Annual artillery pieces, United Kingdom and Austria-Hungary, 1914–1918; Ellis & Cox, The World War I Databook, as reproduced by WorldWars.net. These are guns, not shells. The comparison does not measure calibre, effectiveness, the whole Allied/Central industrial base or total munitions output.
The factory behind the front
Horace Nicholls photographed workers among rows of shells at the National Filling Factory at Chilwell in July 1917. The scene makes the scale of wartime supply tangible: a weapon at the front depended on manufacturing, filling, inspection, storage and transport far behind it. Men and women appear together in an industrial environment shaped by the demand for ammunition. The photograph complements the artillery-production chart without depicting the same statistical unit: the chart counts guns, while the warehouse contains shells. It is a reminder that a complete picture of military production requires several linked measures, rather than one undifferentiated total of “munitions”.
Photo: Horace Nicholls. Chilwell shell warehouse, July 1917; IWM Q 30011. Public domain (expired Crown copyright). Displayed smaller; uncropped.
6. Casualties in major Western Front battles
The cost of a battle depends partly on how its boundaries are drawn. A short engagement, a months-long offensive and a campaign containing several battles are not automatically comparable. This chart therefore ranks a selected set of major Western Front engagements and states the scope behind each estimate. It is not a definitive ranking of the whole world war. Casualties include people killed or wounded and, in some source totals, those missing or captured; they are not all deaths. Several figures are rounded lower benchmarks, while the Passchendaele estimate is shown with an extension for the published range. Differences between sources are meaningful evidence about the limits of counting, rather than a reason to supply extra decimal places. Select an engagement to read its dates and the components behind the plotted total. The bars help communicate scale, but they cannot by themselves establish whether an operation achieved its objectives.
Sources: National Army Museum: 1916; Verdun Memorial: battle losses; Imperial War Museums: Passchendaele fact sheet; National Army Museum: 1917; National Army Museum: 1918; National Army Museum: 1915. Seven selected Western Front engagements, not a definitive worldwide ranking. Figures combine opposing forces. Definitions differ: Verdun and the cited Somme estimate cover killed and wounded; other totals may include missing and prisoners. Somme and Arras are lower benchmarks. The dashed extension for Third Ypres shows the cited upper estimate. Inspect each bar for dates, definitions and qualifications.
7. Where British Empire army troops came from
Hover, tap or choose an item for detailed values.
The British imperial war effort drew on populations with very different political relationships to Britain. The War Office’s 1922 summary records more than eight million troops from the British Isles, dominions, India and other colonies. The treemap keeps the original categories so that the areas reconcile to the source’s stated total. It is an army-troop comparison, not a complete census of everyone who contributed to the war. Naval personnel, merchant seafarers and many labour services require separate accounting, and the figures do not mean that every person served overseas. Nor should service be read as evidence of a single shared motivation: enlistment and participation occurred within different systems of recruitment, obligation and imperial power. India’s category refers to the colonial-era territory and institutions, rather than the boundaries of the present-day republic. Choosing a rectangle reveals its count and share within this particular historical table.
Sources: War Office, Statistics of the Military Effort of the British Empire (1922), p.756. War Office (1922), printed page 756: 8,586,202 British, Dominion, Indian and Colonial army troops. Categories retain historical source geography. Includes troops already serving and wartime recruits; not all served overseas. This army-only series differs from the all-services mobilisation estimate in chart 2 and must not be added to it.
Indian troops on the Western Front
This photograph from H. D. Girdwood’s collection shows Indian infantry at Fauquissart in France in August 1915, prepared against a gas attack. It gives a human setting to the imperial troop categories in the treemap. The British Library notes that the collection was produced for the Government of India and that some scenes may be posed reconstructions rather than photographs taken during combat. That context matters: an official photograph can document people, equipment and military presentation without being an unmediated record of a particular attack. The men pictured should not be treated as a statistical sample of the Indian Army’s wider wartime service.
Photo: H. D. Girdwood / British Library. Indian infantry at Fauquissart, 9 August 1915; Photo 24/(300). No known copyright restrictions (Flickr Commons). Displayed smaller; uncropped.
8. National debt before and after the war
The financial burden of war lasted beyond the fighting. These connected dots compare government debt relative to national output in 1913 and 1920 for five countries with both observations in the selected research table. The United Kingdom, France, Italy and Belgium were belligerents; the neutral Netherlands provides a useful comparison rather than another example of Allied borrowing. Measuring debt as a percentage of GDP avoids putting unlike national currencies on the same numerical scale. It also means that the ratio can change because of movements in debt, output or prices. The comparison includes the early post-war years and cannot attribute every change solely to wartime military expenditure. Countries with missing post-war observations, including states whose political boundaries changed dramatically, are excluded rather than assigned zero. The length of each connector shows the change in percentage points, while the two dots preserve the starting and ending levels.
Sources: Lopez & Mitchener (2018), Table 1. Public debt / GDP in 1913 and 1920, selected countries with paired observations in Lopez & Mitchener, Table 1. Netherlands is a neutral comparison. Changes include post-war borrowing, inflation and GDP movements; they are not a direct measure of war expenditure. Missing countries are omitted rather than estimated.
9. Military deaths and civilian death estimates
Military and civilian deaths describe overlapping historical experiences but distinct counting categories. This chart uses four countries from an Open University comparison to show the military estimate alongside the lower published civilian estimate, with a lighter extension to the upper civilian figure. That extension is a spread across differently scoped estimates, not a confidence interval or a count of additional identified victims. Civilian totals can change greatly depending on the treatment of hunger, disease, occupation, displacement and deaths after the armistice. The chart does not add the worldwide influenza death toll to wartime losses. Its military estimates also come from a different scholarly compilation from the opening map, so differences between those displays are retained and explained rather than silently harmonised. These selected countries do not represent a complete global toll: losses elsewhere, including the Ottoman Empire and colonial Africa, remain outside the comparison.
Sources: Open University: military and civilian losses. Four selected country examples from the Open University comparison. Civilian ranges reflect differing published estimates and coverage, not statistical confidence intervals. France’s upper estimate is particularly disputed. The light extension is the difference between high and low estimates, not another known category of deaths. Military estimates come from a different compilation than chart 1. No worldwide total is inferred; Ottoman and colonial civilian losses are not represented.
10. When countries entered and left the war
Hover, tap or choose an item for detailed values.
Countries did not all enter the war together, and they did not all stop fighting on the same date. This timeline distinguishes the initial belligerents from later entrants such as Italy, Portugal and the United States. Most endpoints mark an armistice or the end of the principal fighting, rather than a final peace treaty. Russia’s bar ends at Brest-Litovsk, while Romania’s interrupted participation is drawn as two separate segments. The tiny final Romanian segment marks its return immediately before the German armistice. A long bar does not mean continuous combat everywhere: occupation, exile, changing fronts and the time needed to deploy forces could alter a country’s practical involvement. The British Empire is represented by Britain’s entry date, although individual contingents reached different theatres later. Select a country for its precise date convention and qualifications. Read this political timeline alongside the campaign timeline, which answers a different question about where fighting occurred.
Sources: National Army Museum timeline; Library of Congress: declarations of war. Dates describe selected countries’ belligerency, using entry into hostilities or declarations and effective armistice/peace endpoints. They do not mean continuous fighting throughout. Romania’s December 1917 armistice and November 1918 re-entry appear as separate segments. Montenegro continued as a belligerent through exile despite its army’s 1916 capitulation. Two-pixel minimum makes very short segments selectable.
Major belligerents: reference table
Selected belligerents, with mobilisation estimates from the 1924 War Department series reproduced by Britannica and population/military deaths from Broadberry & Harrison. Historical territorial units and source coverage differ. A dash means no separate comparable value in the chosen source, not zero. Search any column or click a heading to sort. Dates and exceptions follow chart 10.
16 countries
| Austria-Hungary | Central Powers | 1914-07-28 | 1918-11-04 | 58,600,000 | 7,800,000 | 1,100,000 | Armistice effective 4 November; empire disintegrating. |
|---|---|---|---|---|---|---|---|
| Serbia | Allied | 1914-07-28 | 1918-11-11 | — | 707,343 | — | Army continued the war after the 1915 invasion and retreat. Published population and military-death comparison combines Serbia and Montenegro: 4.9m people and 278,000 deaths; separate values are not inferred. |
| Russia | Allied | 1914-08-01 | 1918-03-03 | 167,000,000 | 12,000,000 | 1,811,000 | Brest-Litovsk peace; an armistice had begun on 15 December 1917. |
| Germany | Central Powers | 1914-08-01 | 1918-11-11 | 67,800,000 | 11,000,000 | 2,037,000 | German armistice. |
| France | Allied | 1914-08-03 | 1918-11-11 | 39,600,000 | 8,410,000 | 1,327,000 | German declaration of war to German armistice. |
| Belgium | Allied | 1914-08-04 | 1918-11-11 | 7,600,000 | 267,000 | 38,000 | German invasion to German armistice. |
| British Empire | Allied | 1914-08-04 | 1918-11-11 | 388,300,000 | 8,904,467 | 913,000 | British declaration also involved the empire; units deployed at different dates. Population/deaths combine the UK and empire rows in Broadberry–Harrison; the treemap measures army troops only. |
| Montenegro | Allied | 1914-08-05 | 1918-11-11 | — | 50,000 | — | Military capitulation in January 1916; government remained in exile. Published population and military-death comparison combines Serbia and Montenegro: 4.9m people and 278,000 deaths; separate values are not inferred. |
| Japan | Allied | 1914-08-23 | 1918-11-11 | — | 800,000 | — | Declaration against Germany; bar does not imply continuous front-line combat. No matching population/death row in the chosen comparison; mobilisation retained. |
| Ottoman Empire | Central Powers | 1914-11-02 | 1918-10-31 | 21,700,000 | 2,850,000 | 804,000 | Russian declaration after the Ottoman naval attack on 29 October; Mudros armistice effective 31 October. |
| Italy | Allied | 1915-05-23 | 1918-11-04 | 35,900,000 | 5,615,000 | 578,000 | War against Austria-Hungary; declared against Germany in August 1916. |
| Bulgaria | Central Powers | 1915-10-14 | 1918-09-30 | 4,700,000 | 1,200,000 | 88,000 | War against Serbia; Salonika armistice effective 30 September. |
| Portugal | Allied | 1916-03-09 | 1918-11-11 | 6,100,000 | 100,000 | 7,000 | German declaration; colonial clashes had occurred earlier. |
| Romania | Allied | 1916-08-27 | 1917-12-09 | 7,600,000 | 750,000 | 250,000 | Focșani armistice; peace followed in May 1918. Re-entered 10 November 1918. |
| United States | Allied | 1917-04-06 | 1918-11-11 | 98,800,000 | 4,355,000 | 114,000 | Associated Power; declaration against Germany. |
| Greece | Allied | 1917-06-30 | 1918-11-11 | 4,900,000 | 230,000 | 26,000 | Unified government’s formal entry after the National Schism. |
Other pages
Sources and image credits
Sources: Broadberry & Harrison, The Economics of World War I, Table 9; Britannica: mobilisation and losses; National Army Museum timeline; Archibald Hurd, The Merchant Navy, Volume III, Appendix C; Ellis & Cox production figures, reproduced by WorldWars.net; National Army Museum: 1916; Verdun Memorial: battle losses; Imperial War Museums: Passchendaele fact sheet; National Army Museum: 1917; National Army Museum: 1918; National Army Museum: 1915; War Office, Statistics of the Military Effort of the British Empire (1922), p.756; Lopez & Mitchener (2018), Table 1; Open University: military and civilian losses; Historical basemaps: approximate 1914 borders (GPL-3.0); Library of Congress: declarations of war. Calculations and visualisations by The World in Stats. Body photographs are credited beside their images. Banner: 1st Australian Division near Harbonnières, 9 August 1918; IWM E (AUS) 2878 — Australian official photographer, via Wikimedia Commons; Public domain (expired Crown copyright).