Shared electric scooters and electric bicycles have become common features of urban mobility in the United States since the first dockless scooter fleets appeared in 2017. Cities adopted these devices as instruments for short trips, first- and last-mile connections, congestion relief, and lower-carbon travel, and their use has grown in parallel with public investment in cycling infrastructure and dedicated lanes.

By the early 2020s, shared and privately owned e-scooters accounted for tens of millions of trips annually in the largest metropolitan areas, and several states reported e-bike purchases rising faster than any other vehicle category. Electric bicycle sales, in particular, expanded sharply during and after 2020, outpacing conventional bicycle sales in several national markets and attracting riders who had not previously cycled for transport. This rapid diffusion has not been free of cost.

Emergency department surveillance shows a rising volume of injuries associated with these devices, and national analyses now treat micromobility trauma as an emerging public health problem rather than an isolated urban novelty.( 1,2) Unlike conventional bicycles, e-scooters expose riders to small wheels, a standing posture, and forward ejection over the handlebars, often without helmets, while e-bikes combine bicycle ergonomics with motorized acceleration, higher mass, and greater speed.( 3,4)

These mechanical differences imply that the two devices may produce different injury profiles, yet they are frequently grouped together in policy and surveillance reporting, which can obscure where prevention should be directed. The regulatory response has been similarly fragmented: helmet mandates, speed limits, and age restrictions vary across states and cities, and no single national standard governs the design or operation of either device. Reliable national estimates of injury burden are a prerequisite for that regulation, because they establish the size and composition of the problem that policy must address.

Available estimates anticipate a steep upward pattern for both devices. National surveillance identified roughly 4,600 electric scooter injuries in 2014 and about 29,000 in 2019, a rise that coincided with the rollout of shared fleets.( 1)

Federal reports from the Consumer Product Safety Commission place micromobility injuries in the tens of thousands per year by 2021 and note that e-scooter and e-bike cases carry a meaningful share of fractures and hospitalizations.( 4)

A national comparison of electric and conventional scooters and bicycles found that electric devices carry a higher risk of hospitalization than their non-motorized counterparts.( 5)

Subsequent analyses of the same system report that electric bicycle injuries have increased faster than injuries from conventional bicycles, with fractures and hospital admissions rising most among adults.( 6,7)

Head and neck injuries from powered scooters and e-bikes follow a similar trajectory, and they concentrate in young men for scooters and in older riders for e-bikes.( 8,9)

Pediatric series add that fractures from powered scooters have climbed among children and adolescents even as conventional scooter injuries have stabilized.( 10)

Regional surveillance in Illinois and state-level studies confirm the same direction at subnational scales.( 11) The expected pattern, therefore, is one of accelerating national counts for both devices, with e-scooter injuries skewed toward younger adults and e-bike injuries toward middle-aged and older riders.

Most national studies have examined a single device, a shorter window, or a period ending before 2023, so the comparative trajectory of e-scooter and e-bike injuries across the full adoption period remains uncertain.( 2,12)

The years after 2020 are particularly informative because they capture the pandemic shift toward e-bike commuting and the return of shared scooter fleets, yet they are the least represented in the published record. A direct comparison of the two devices over the same recent interval, with consistent case definitions and design-based national estimates, would clarify whether prevention resources should be allocated differently to each device.

This study aimed to estimate national counts of e-scooter and e-bike emergency department visits in the United States from 2017 through 2024, to describe their trajectory, to characterize the injuries and age groups that predominate for each device, and to compare the likelihood of hospital admission between them.

The sample contained 12,754 e-scooter and 7,141 e-bike injury visits between 2017 and 2024, and 59 additional records that met both definitions were excluded from the device-specific comparisons. The national estimate of e-scooter injuries climbed from 18,644 in 2017 to 126,420 in 2024, a rise of 578% (Figure 1). The estimate reached 28,450 in 2018 and 43,391 in 2019, and the 2019 value had nearly tripled by 2024.

The trajectory paused in 2020, when the estimate of 40,968 fell below the 2019 value, and then resumed growth through 59,775 in 2021, 66,099 in 2022, and 72,378 in 2023 (Table 1). The e-bike estimate rose 636% over the same period, from 12,015 in 2017 to 88,429 in 2024. The e-bike curve remained near 11,757 in 2018, reached 16,776 in 2019, and then accelerated to 24,615 in 2020, 30,558 in 2021, 46,262 in 2022, and 54,663 in 2023.

The average annual change was 9,646 visits for e-scooters (p = <0.001) and 5,876 visits for e-bikes (p = 0.006). Across the full period, the pooled estimate was 456,126 e-scooter visits (95% CI 392,462-519,789) and 285,075 e-bike visits (95% CI 249,662-320,489). The e-bike estimate exceeded the e-scooter estimate in no single year, although the gap narrowed from 2022 onward. The confidence intervals for 2024 were wider than for the preceding years (Table 1).

The age distribution differed between devices (Figure 2).

E-scooter injuries clustered in young adults, with 35.0% of cases among patients aged 18 to 34 years, whereas e-bike injuries spread more evenly across adulthood and peaked in patients aged 55 years or older at 28.4% (Table 2).

The modal age group was 18 to 34 years for e-scooters and 55 years or older for e-bikes. Children and adolescents accounted for 22.7% of e-scooter and 24.3% of e-bike cases, and adults aged 35 to 54 years for 22.8% and 23.7%. The median age was 27 years for e-scooter riders and 31 years for e-bike riders. Males predominated in both groups, at 63.3% of e-scooter and 71.5% of e-bike cases, and females represented 36.6% and 28.4%; the male predominance was stronger for e-bikes, and the female share of e-scooter injuries was the larger of the two.

Fractures were the leading diagnosis in both devices, at 27.0% of e-scooter and 22.1% of e-bike injuries (Figure 3). Contusions and abrasions accounted for 19.0% and 17.1%, and lacerations for 13.0% and 14.0% (Table 2). Strains and sprains were less frequent, at 6.7% and 7.3%. Internal organ injuries accounted for 11.7% and 11.4% of cases, and concussions for 2.4% and 1.8%. The residual diagnosis category was the second largest for e-bikes, at 26.3%, and accounted for 20.3% of e-scooter cases.

The distribution of body regions mirrored the diagnosis pattern (Figure 4). Head, face, or neck injuries accounted for 33.1% of e-scooter and 29.0% of e-bike cases (Table 2). Upper extremity injuries were present in 28.2% and 23.5%, and lower extremity injuries in 26.0% and 26.3%. Trunk injuries were more frequent in e-bike cases, at 18.0% versus 11.1%. The remaining body regions accounted for 1.5% and 3.2%.

The proportion hospitalized was 11.0% for e-scooters and 12.0% for e-bikes, and the proportion treated and released was 84.0% and 83.0% (Table 2). The remaining cases were transferred, observed, or left without being seen. The difference in hospitalization of −1.1 percentage points (95% CI −2.8 to 0.6) was not statistically significant (p = 0.218).

The sensitivity analysis restricted to narrative text without product codes produced a pooled e-scooter estimate of 88,831 visits, and the annual series retained the same direction and growth as the primary definition. This estimate was lower than the primary pooled estimate, as expected from the removal of code-identified cases.

The main finding is a steep and sustained rise in emergency department injuries from both e-scooters and e-bikes between 2017 and 2024, with e-scooters contributing more visits overall and e-bikes growing at a slightly faster relative pace. The two devices also produced distinct profiles: e-scooter injuries concentrated in young adult men with fractures and head and neck injuries, whereas e-bike injuries spread across a broader age range and carried a similar likelihood of hospital admission.

The growth pattern is consistent with earlier national surveillance. The estimated incidence of electric scooter injuries in the United States rose from 2014 to 2019 and was attributed to the expansion of shared fleets,( 1) and national analyses through 2022 found a continuing increase in powered scooter and e-bike injuries.( 2,15)

Single-city studies of dockless rental scooters in large United States metropolitan areas documented the same influx of injured riders soon after fleet launch, with young adult men overrepresented.( 16,17)

Orthopaedic series from level I trauma centers likewise described a rapid rise in e-scooter fractures as the devices became available.( 18,19)

The temporary decline in e-scooter visits observed in 2020 coincides with pandemic mobility restrictions and the interruption of shared services, a pattern also reported for fracture injuries associated with alternative transport during COVID-19.( 20)

The uninterrupted rise of e-bike injuries across the same period is consistent with the pandemic surge in e-bike sales and use documented in national data.( 6,7)

International evidence from China, where e-bikes have been established for decades, shows that e-bike crashes are a leading source of two-wheeled road injury, and systematic reviews there link higher speeds and mixed traffic to more severe outcomes than pedal bicycles.( 21)

Comparisons from Europe and Australia describe the same secular increase in e-scooter trauma after fleet introductions, which suggests that the United States pattern reflects a general property of device adoption rather than a local anomaly.( 22,23)

Oslo, Helsinki, and several German cities reported steep increases in e-scooter and e-bike emergency presentations after the arrival of shared fleets, with injury severity that closely resembles the present estimates.( 24-26)

The magnitude of the burden is consistent with the federal estimates that anchor this literature. The Consumer Product Safety Commission reported that e-scooter and e-bike injuries together reached tens of thousands of emergency department visits per year by 2021, and the present pooled estimate of more than 700,000 visits across eight years falls within the same order of magnitude while extending the series to 2024.( 4)

The difference in totals between devices is also consistent with recent national comparisons, which find more scooter-related visits than e-bike visits but a faster growth rate for e-bikes.( 2,15)

These correspondences increase confidence that the narrative-based case definition captured the intended population despite the absence of a single consistent product code across the period, and they imply that the annual totals reported here can be interpreted as national burdens rather than as idiosyncratic samples.

Comparability with earlier estimates is limited by differences in case definitions. Earlier estimates used narrative terms alone, product codes alone, or a combination of the two, which produces different point estimates for the same year even when the underlying data are identical.( 2,15)

The present approach of combining narrative and product codes, while excluding non-electric lookalikes, was intended to maximize sensitivity in the early years and specificity throughout, and the close agreement with the Consumer Product Safety Commission figures for overlapping years supports this choice.( 4)

Readers comparing these numbers with single-city or single-year series should account for those definitional differences, which affect the level of each estimate more than the shape of the trend.

The age distributions align with prior work but sharpen the contrast between devices. The clustering of e-scooter injuries in young adults matches single-center and multicenter series that place most injured riders between 18 and 34 years.( 27,28)

The older profile of e-bike riders, with more than one quarter aged 55 years or older, is consistent with national age-specific analyses showing that e-bike injury growth is fastest in middle-aged and older adults.( 6,29)

Sex distributions also agree with the literature, which uniformly reports male predominance for both devices, usually above 60%.( 30,31)

The greater male predominance for e-bikes than for e-scooters is consistent with the demographic composition of e-bike riders described in national samples.(6) The pediatric contribution to e-scooter injuries, at about one fifth of cases, is in line with pediatric emergency series that document a rising share of powered scooter injuries among children and adolescents, and with guidance that frames micromobility injuries as a growing pediatric injury mechanism.(10,32,33) Because many shared platforms set a minimum rider age, this pediatric burden may reflec

t private devices or noncompliance with rental terms, which points to parental supervision and device design as relevant levers in addition to enforcement.

The injury patterns replicate the established signature of micromobility trauma. Fractures accounted for roughly one quarter of cases, close to the proportions reported in European trauma registries and systematic reviews.( 34,35)

Head and neck involvement in about one third of e-scooter injuries is the finding with the clearest prevention implication, and it is consistent with the high prevalence of facial and cranial injuries described in United States and international series.( 9,36,37) Imaging-based studies of computed tomography findings reinforce the point, showing a substantial share of e-scooter crash patients with intracranial or facial fractures even among those without a recorded concussion.( 9)

The elevated proportion of trunk injuries among e-bike riders has also been reported and may reflect different crash kinematics, including greater speed and collisions with motor vehicles.( 38,39)

Concussion and other head injuries from electric biking are rising nationally, which reinforces the interpretation that e-bike crashes are not simply bicycle crashes with a motor.( 8,40)

The similarity in hospital admission between devices, around 11% to 12%, stands out given the different age profiles. Comparable admission proportions have been reported in multicenter e-scooter and bicycle comparisons, although some trauma-center series describe higher rates because they capture only the most severely injured patients.( 35,41)

The finding that e-bike riders are hospitalized at least as often as e-scooter riders, despite being older, supports concerns that e-bike crashes can be high-energy events.( 37,40)

Cost studies add that the burden extends beyond the index visit, with orthopaedic injuries from both devices generating substantial follow-up and surgical care.( 34,42)

Trauma-center comparisons of injury severity add nuance to the present disposition-based picture. Several European registries report higher Injury Severity Scores for e-bike than for e-scooter crashes, driven by craniocerebral and truncal trauma, even though e-scooter series often include more patients overall.( 37,39)

This is compatible with the observation here that admission proportions were similar despite very different age structures, and it suggests that the average e-bike crash may be more energy-intensive than the average e-scooter crash. Severity data are not available in NEISS, which is why this study relied on disposition as a proxy for serious injury, and this proxy probably underestimates the full spectrum of severe outcomes.( 40)

Mechanistic explanations for these patterns are plausible and supported by the literature. E-scooter riders are exposed to small wheels, a standing posture, and forward ejection over the handlebars, which concentrates impact on the face, head, and upper extremities.( 3,43)

E-bike riders travel at higher speeds and share roadways with motor vehicles, which raises the frequency of high-energy and truncal injuries.( 38)

The older age of e-bike riders may also contribute through reduced bone resilience and slower reaction times, mechanisms proposed in age-specific analyses of e-bike trauma.( 29)

Helmet use remains low in both groups, and alcohol involvement is repeatedly documented in e-scooter crashes.( 44,45)

Helmet status can now be extracted from unstructured clinical notes at scale, which offers a route to quantify the protective effect in future surveillance.( 43)

Behavioral factors also differ between the two populations and may shape prevention. Alcohol and nighttime riding are overrepresented in e-scooter injuries, and several cities have responded with nighttime restrictions or speed limits that were followed by measurable reductions in injuries.( 44,45)

E-bike injuries, by contrast, are more often associated with traffic exposure and roadway interactions than with intoxication, which shifts the preventive emphasis toward infrastructure and vehicle design.( 38,40)

These distinctions reinforce the case for keeping the two devices separate in injury surveillance and in safety policy.

Helmet use deserves particular emphasis because head and face injuries were the single largest body-region group. Observational series consistently associate helmet nonuse with more severe cranial and facial injuries in e-scooter crashes, and helmet legislation has been proposed as the most direct lever for reducing that burden.( 9,37) A

t the same time, helmet enforcement is difficult for shared fleets, which may explain why helmet prevalence has remained low despite repeated recommendations.( 44)

The higher trunk injury share in e-bike riders suggests that helmets alone will not address the full injury burden of that device, and that speed management and separation from motor traffic deserve equal attention.( 38,39)

This study has several strengths. It used a nationally representative probability sample with design-based variance, covered the full adoption period through 2024, and applied an explicit, reproducible case definition that handles the 2020 product code change. Several limitations qualify the estimates. Case identification relied partly on free-text narratives, which can misclassify devices with sparse descriptions and probably under-ascertain early cases, a limitation shared with other narrative-based NEISS analyses.( 15)

The sensitivity analysis restricted to narrative text confirmed the direction of the trend but produced lower totals, which indicates that product codes capture cases the narrative alone misses. The 2024 file contained records without assigned sample weights that could not contribute to national estimates, and the estimates therefore understate the most recent year to an unknown degree. NEISS records emergency department visits only and provides no trip or distance denominators, so the estimates measure burden rather than risk per trip, and comparisons between devices cannot be read as comparisons of per-ride risk.( 46)

The data also exclude injuries treated in urgent care or primary care, deaths that occurred without an emergency department visit, and exposures among tourists or visitors whose visits fall outside the sampling frame. Finally, the data contain limited information on helmet use, alcohol, and crash circumstances, which prevents a fuller analysis of causation. Future studies could link these records to exposure data, incorporate helmet and speed information, and validate the case definition against state trauma registries.

The implications are device-specific. For e-scooters, the concentration of injuries in young adults and the head and neck burden argue for helmet promotion, reduced speeds, and continued limits on intoxicated riding.( 44,47)

For e-bikes, the older rider population and the growing fracture burden suggest attention to vehicle stability, roadway infrastructure, and training for riders who may be returning to cycling after a long gap.( 7,21)

These two sets of measures share little beyond their target population, and their success should be judged separately and transparently over time. Emergency physicians and trauma surgeons can contribute by documenting the device type in the medical record and by reinforcing helmet counseling at discharge, which in turn improves the surveillance that these estimates depend on.( 43)

Regulation and infrastructure designed for one device will not automatically serve the other, and surveillance systems that distinguish e-scooters from e-bikes will be needed to evaluate the effect of any intervention.( 46,47)

Taken together, these results extend the national surveillance record to the end of 2024 and show that e-scooters and e-bikes, although often treated as a single micromobility category, generate different injury problems that call for different countermeasures. The estimates provide a national reference against which state and local programs can benchmark their own data, and they identify the age groups and injury types most likely to benefit from intervention in each device population. As adoption of both devices continues, periodic updating of these estimates will indicate whether prevention efforts are keeping pace with the growth in use.

The national estimates of e-scooter and e-bike injuries treated in United States emergency departments increased sharply between 2017 and 2024, with e-scooters contributing more visits overall and e-bikes growing at a faster relative pace. E-scooter injuries concentrated in young adult men and were dominated by fractures and head and neck injuries. E-bike injuries spread across older age groups and produced a comparable rate of hospital admission. The device-specific age and injury profiles identified here support targeted prevention rather than uniform micromobility policy.