How have civilian casualty trends in Afghanistan changed over time and after major events like the 2021 US withdrawal?

Checked on February 7, 2026
Disclaimer: Factually can make mistakes. Please verify important information or breaking news. Learn more.

Executive summary

Civilian casualties in Afghanistan rose and fell in patterns tied to shifts in tactics and foreign involvement: airstrike-related deaths surged in the late 2010s, insurgent use of IEDs and ground fighting drove sharp increases in 2020–2021, and a dramatic spike in documented civilian harm occurred as US and NATO forces withdrew in mid‑2021 [1] [2] [3]. United Nations monitoring recorded a 47% rise in civilian killed and wounded in the first half of 2021 versus 2020 and warned that 2021 was on track to be the deadliest year since systematic records began in 2009 [3] [4].

1. A decade of changing drivers: airpower, IEDs, and shifting blame

Over the 2010s the sources of civilian harm in Afghanistan shifted: international airstrikes were a growing cause of civilian deaths by 2017–2019, with UN and monitoring groups reporting international airstrike casualties more than tripled between 2017 and 2019 and a large share of child deaths tied to such strikes [1] [5]. At the same time insurgent tactics—improvised explosive devices (IEDs), targeted killings and ground engagements—continued to inflict large numbers of casualties, meaning the pattern of who was inflicting harm oscillated with the operational mix on the ground [2].

2. The 2020–2021 transition: negotiations, withdrawal and a grim uptick

The US–Taliban agreement and subsequent drawdown altered the battlefield calculus: while peace talks began earlier, the period around the formal withdrawal saw intensified Taliban offensives and expanded fighting between Afghan actors, and UN monitoring recorded a marked rise in civilian harm beginning in May 2021 as international troops departed and the Taliban seized territory [6] [2] [7]. UNAMA’s mid‑year update documented 5,183 civilian casualties in the first six months of 2021 (1,659 killed and 3,524 injured), a 47% increase from the same period in 2020, and pointed to May–June as the period of sharpest escalation [3] [4].

3. Who was responsible and who suffered most

UN monitors attributed the majority of civilian casualties in that period to anti‑government forces—about 64%—with the Taliban and Islamic State‑K among those cited, and emphasized the disproportionate toll on women and children (children made up about 32% of casualties in the mid‑year 2021 report) [8] [9]. The UN also noted that for the first half of 2021 no civilian casualties were attributed to international military action, reflecting the changing balance of force after the withdrawal began [4].

4. Humanitarian and medical consequences as violence spread

Medical responders documented the operational consequences: Doctors Without Borders reported that violence after the withdrawal spread into cities and frontlines around hospitals, limiting access to care and increasing the number of patients with war wounds treated since May 2021, underscoring how spikes in fighting translate quickly into strain on health services and higher civilian mortality and morbidity [10]. UNAMA similarly warned that without de‑escalation Afghanistan risked its highest annual civilian toll since monitoring began [3].

5. Limits of the record and how to interpret trends

All authoritative sources caution the record is incomplete: UNAMA explicitly acknowledges potential under‑reporting given operating constraints and the disruptions of COVID and active conflict, and military reporting practices (for example US CENTCOM’s cessation of published monthly airstrike data in 2020) create data gaps that complicate comparisons across years [3] [1]. Reporting shows clear inflection points—rising airstrike casualties late in the decade, rising insurgent‑caused casualties into 2021, and a pronounced spike coincident with the international withdrawal—but exact totals and attribution remain conservative estimates constrained by access and data gaps [1] [3].

Want to dive deeper?
How did the geographic distribution of civilian casualties in Afghanistan shift before and after the 2021 withdrawal?
What role did the Afghan Air Force play in civilian casualties in 2019–2021 and how was it trained and equipped by international partners?
How reliable are UNAMA casualty attribution methods and what are the main sources of under‑reporting in Afghanistan?