Propeller injuries: what each study found, kept apart.

How to read these figures

• Each study stands alone. The studies use different years, different age groups and different methods, so their figures are never added together or averaged here.
• Emergency-department counts are not risk. The US studies use the National Electronic Injury Surveillance System (NEISS), run by the Consumer Product Safety Commission. NEISS records visits to hospital emergency departments. Its figures describe who arrived at hospital and with what, not how likely any pilot is to be hurt.
• Drones are hard to find in NEISS. NEISS has no product code for drones. Researchers find drone cases by searching the written case narratives under code 5021, toy vehicles (excluding riding toys). According to the University of Florida publication below, NEISS draws on about 100 hospitals chosen as a probability sample of the more than 5,000 US hospitals with emergency departments.
• No study is about racing. None of the studies separates racing drones from other drones, and none separates propeller contact from other kinds of impact. The only racing-specific item found is a single case report.

University of Florida IFAS, publication AE560 (all ages)

Gorucu and Ampatzidis analysed NEISS patients treated from 1 January 2015 to 31 December 2019, of all ages, found with the search term "drone". They had 89 cases, which NEISS weighting turns into national estimates. The copy read for this site says the publication was revised in November 2024 and February 2026.

The authors sum up the typical case as an adult man with cut fingers. Case narratives quoted in the publication include a man whose finger was caught in the blades while playing with a drone, and a 21-month-old child struck in the eye by a drone. The dataset includes one death, by electrocution, involving a metal pole and a drone caught on power lines; it was not a propeller injury.

The national total disagrees with itself

This publication does not give one consistent national figure, so none is quoted here as settled. Its results give approximately 4,250 injuries for 2015 to 2020, with a 95 percent confidence interval of 3,625 to 3,790, a range that does not contain 4,250. Its method section covers 2015 to 2019, and its summary says more than 3,700. All three versions are set out here because the source itself does not reconcile them.

The authors write that propeller guards could create a safe zone and give some protection from propeller injuries. That is their recommendation. The study did not measure it.

Khan and Brown, Cureus, 2021 (children)

This study searched NEISS for 2010 to 2019, patients aged 0 to 18. From 3,230 toy-vehicle records the authors found 26 drone-related cases. They made no national estimate.

• No drone injuries appear in the sample for 2010 to 2014; yearly counts rose from 1 in 2015 to 9 in 2018.
• Head and face injuries were 58 percent (15 of 26). Lacerations were the most common diagnosis, at 42 percent (11 of 26).
• The study splits direct impact by the drone, 65 percent (17 cases), from indirect injuries, 35 percent (9 cases), such as retrieving a drone from a tree or roof, chasing one, swallowing a battery or exposure to smoke. It does not separate propeller contact from other impacts.
• 92 percent of the 26 children were treated and discharged. Two were admitted, one with a cut to the cornea.
• The paper disagrees with itself on sex: its table gives 9 of 26 (35 percent) female, its text 42 percent female.

The authors state that their sample is small, that NEISS case descriptions are limited, and that coding errors may misstate the true number.

Johnson, Svach and Brown, 2019 (not read)

A 2021 case report cites a 2019 paper in the American Journal of Preventive Medicine, "Drone and other hobbyist aircraft injuries seen in U.S. emergency departments, 2010-2017", as estimating about 12,842 hobbyist drone injuries treated in US emergency departments over those years. The paper itself was not read for this site, so the figure is unconfirmed and rests only on that citation.

Case reports

A case report describes one patient. It shows that an injury happened, not how often.

Crahay, Rampat, Tonglet and Rakic, BMJ Case Reports, 2021

A 60-year-old man who flew racing drones was struck in the face by his own four-propeller racing drone as it came in to land, injuring his face, eyelid and eye. The authors call it the first reported adult case of eye injury from a drone. They report that he had not kept a safe distance, and they encourage protective eyewear and better drone design, suggesting a helmet or safety goggles for pilots and people nearby. Those are the authors' suggestions; the report does not test them.

Bansal, Aggarwal and Tiwari, Journal of Clinical Orthopaedics and Trauma, 2021

A 40-year-old man in India had cuts to the fingers of both hands, with two fingertip fractures, from a drone's front propeller blades during landing. The drone was used for municipal surveillance and mapping, not racing. The authors say people handling drones are prone to injury during landing and recommend eye protection, gloves and propeller guards; this is their recommendation, not a tested finding. Their review also cites two studies not read for this site: Duma and colleagues (2021), reported as finding that blade tip speed correlated most strongly with injury, and Campolettano and colleagues (2017), reported as linking severity to speed, height and drone mass.

Moskowitz and colleagues, American Journal of Ophthalmology Case Reports, 2018

Seen only as a search summary: a case of a child whose eye was ruptured by a remote-controlled drone's propeller. The article was not opened, so the details are unconfirmed.

FAA-sponsored research on injury potential

The FAA's Center of Excellence for unmanned aircraft (ASSURE) published Task A4, "UAS Ground Collision Severity Evaluation, Revision 2", in 2017 (Arterburn, Ewing, Prabhu, Zhu and Francis). It studied three kinds of injury from small drones: blunt force trauma, penetration and laceration. Its abstract names the drone characteristic most significant for each: kinetic energy at worst-case terminal velocity or maximum cruise airspeed for blunt force trauma, energy density for penetration, and rotor diameter for laceration. It is engineering research on collision severity, not a count of injuries.

Propellers in the racing rules

Racing bodies and associations set propeller rules, each within its own framework. The FAI Sporting Code, Volume F9, 2025 edition (effective 1 May 2025), rule C.1.3, limits F9U propellers to 6 inches (15.2 cm) and forbids full metal propellers. The same edition's change log records that the earlier ban on propeller protection devices, present in the 2019 edition, was removed in changes effective 1 January 2020, stating there is no safety reason to prohibit them. A 2026 edition exists but was not read; its effective date is reported as both 1 January and 1 June 2026. The MAAA's FPV policy MOP066 (19 August 2019), section 9.6, bans carbon fibre and metal propellers in racing multirotors. The two bodies' rules differ and apply only at their own events. More on race safety rules.