NEISS-AI · for journalists

US e-bike injuries: key numbers and free charts

National estimates of e-bike and e-scooter injuries treated in US hospital emergency departments, 2019–2025, from the Consumer Product Safety Commission's NEISS data. Free to use with credit.

Open the live tracker

Key numbers

101,600e-bike injuries treated in US emergency departments in 2025 (95% CI 69,700–133,600)

+71%increase from 2024 (59,300) to 2025, and about 3.6 times the 2022 level (28,000)

About 3×e-bike emergency visits among people 65 and older, from about 3,200 (2022) to 9,700 (2025)

168,800e-scooter injuries treated in US emergency departments in 2025 (95% CI 97,900–239,700)

All numbers are national estimates with 95% confidence intervals. Updated October 2026 from the complete CPSC NEISS files.

Free chart

Bar chart of US e-bike injuries treated in emergency departments: about 5,200 in 2019, 8,200 in 2020, 15,000 in 2021, 28,000 in 2022, 38,000 in 2023, 59,000 in 2024 and 102,000 in 2025, with 95% confidence intervals.

Download chart (PNG)One-page fact sheet (PDF)

Free to publish with the credit line: “Source: NEISS-AI (Kumar 2026), from CPSC NEISS data.”

Live chart for your website

Paste this into an online story or web page to show an interactive chart that updates when new NEISS data are added. The source and citation are shown under the chart.

<iframe src="https://subash28kumar.github.io/neiss-ai/embed.html?chart=ebike" title="E-bike injuries in US emergency departments (NEISS-AI)" width="100%" height="460" style="border:0;max-width:760px" loading="lazy"></iframe>

Other charts: change chart=ebike to chart=devices (e-bikes and e-scooters) or chart=emoto (regular e-bikes vs e-motos). Preview

How to describe the source

…according to NEISS-AI, a free, open tool by independent researcher Subash Kumar that produces national injury estimates from the Consumer Product Safety Commission's emergency department data.

How it works, in three lines. NEISS is a national sample of about 100 hospital emergency departments run by CPSC. NEISS-AI reads the short injury description in every record with a validated text classifier to identify e-bikes, e-scooters and other devices. It then applies CPSC's survey weights to estimate national totals with 95% confidence intervals.

Accuracy. NEISS-AI's e-bike estimates for 2022–2024 match CPSC's own published figures within 5%. Because it reads the injury descriptions, its counts stay consistent across CPSC's 2024 change in product codes for e-bikes.

Please note.

These are injuries treated in emergency departments, not deaths and not all injuries.
Estimates for 2019–2020 are less precise; please report the range for small numbers.
NEISS-AI is an independent project, not affiliated with or endorsed by CPSC.

About the researcher and contact

Subash Kumar is an independent researcher in Ellicott City, Maryland, who builds open AI tools for road safety and injury prevention. ORCID 0009-0003-6661-3424.

For interviews, custom tables (for example by age group or year) or questions about the data: nivi05subhash@gmail.com. I usually reply within one day.

Cite the software: Kumar S. NEISS-AI: AI-assisted national injury estimates from NEISS emergency department narratives. Version 1.4.0 [software]. Zenodo; 2026. https://doi.org/10.5281/zenodo.23550272